Re: [Qemu-devel] [PATCH] opencores_eth: flush queue whenever can_receive can go from false to true
Il 03/02/2014 05:20, Max Filippov ha scritto:
The following registers control whether MAC can receive frames:
- MODER.RXEN bit that enables/disables receiver;
- TX_BD_NUM register that specifies number of RX descriptors.
Notify QEMU networking core when the MAC is ready to receive frames.
Discard frame and raise BUSY interrupt when the frame arrives but the
current RX descriptor is not empty.
Signed-off-by: Max Filippov jcmvb...@gmail.com
---
hw/net/opencores_eth.c | 33 +++--
1 file changed, 31 insertions(+), 2 deletions(-)
diff --git a/hw/net/opencores_eth.c b/hw/net/opencores_eth.c
index 4118d54..4a44304 100644
--- a/hw/net/opencores_eth.c
+++ b/hw/net/opencores_eth.c
@@ -169,6 +169,7 @@ enum {
};
enum {
+INT_SOURCE_BUSY = 0x10,
INT_SOURCE_RXB = 0x4,
INT_SOURCE_TXB = 0x1,
};
@@ -351,8 +352,7 @@ static int open_eth_can_receive(NetClientState *nc)
OpenEthState *s = qemu_get_nic_opaque(nc);
return GET_REGBIT(s, MODER, RXEN)
-(s-regs[TX_BD_NUM] 0x80)
-(rx_desc(s)-len_flags RXD_E);
+(s-regs[TX_BD_NUM] 0x80);
}
static ssize_t open_eth_receive(NetClientState *nc,
@@ -402,6 +402,12 @@ static ssize_t open_eth_receive(NetClientState *nc,
desc *desc = rx_desc(s);
size_t copy_size = GET_REGBIT(s, MODER, HUGEN) ? 65536 : maxfl;
+if (!(desc-len_flags RXD_E)) {
+open_eth_int_source_write(s,
+s-regs[INT_SOURCE] | INT_SOURCE_BUSY);
+return size;
+}
+
desc-len_flags = ~(RXD_CF | RXD_M | RXD_OR |
RXD_IS | RXD_DN | RXD_TL | RXD_SF | RXD_CRC | RXD_LC);
@@ -551,6 +557,15 @@ static uint64_t open_eth_reg_read(void *opaque,
return v;
}
+static void open_eth_notify_can_receive(OpenEthState *s)
+{
+NetClientState *nc = qemu_get_queue(s-nic);
+
+if (open_eth_can_receive(nc)) {
+qemu_flush_queued_packets(nc);
+}
+}
+
static void open_eth_ro(OpenEthState *s, uint32_t val)
{
}
@@ -567,6 +582,7 @@ static void open_eth_moder_host_write(OpenEthState *s,
uint32_t val)
if (set MODER_RXEN) {
s-rx_desc = s-regs[TX_BD_NUM];
+open_eth_notify_can_receive(s);
}
if (set MODER_TXEN) {
s-tx_desc = 0;
@@ -592,6 +608,18 @@ static void open_eth_int_mask_host_write(OpenEthState *s,
uint32_t val)
s-regs[INT_SOURCE] s-regs[INT_MASK]);
}
+static void open_eth_tx_bd_num_host_write(OpenEthState *s, uint32_t val)
+{
+if (val 0x80) {
+bool enable = s-regs[TX_BD_NUM] == 0x80;
+
+s-regs[TX_BD_NUM] = val;
+if (enable) {
+open_eth_notify_can_receive(s);
+}
+}
+}
+
static void open_eth_mii_command_host_write(OpenEthState *s, uint32_t val)
{
unsigned fiad = GET_REGFIELD(s, MIIADDRESS, FIAD);
@@ -630,6 +658,7 @@ static void open_eth_reg_write(void *opaque,
[MODER] = open_eth_moder_host_write,
[INT_SOURCE] = open_eth_int_source_host_write,
[INT_MASK] = open_eth_int_mask_host_write,
+[TX_BD_NUM] = open_eth_tx_bd_num_host_write,
[MIICOMMAND] = open_eth_mii_command_host_write,
[MIITX_DATA] = open_eth_mii_tx_host_write,
[MIISTATUS] = open_eth_ro,
Reviewed-by: Paolo Bonzini pbonz...@redhat.com
Re: [Qemu-devel] [PULL v4 05/12] milkymist-vgafb: swap pixel data in source buffer
Il 01/02/2014 18:57, Peter Maydell ha scritto: In commit fc97bb5ba3e7239c0b6d24095df6784868dfebbf the lduw_raw() call was eliminated. But we are reading from the target buffer a 16-bit value, which is in big-endian format. Therefore, swap the bytes if we are building for a little-endian host. Paolo, can you remember why you included this change in that commit? It purports to just be moving the display devices around but it seems to have included the introduction of this bug, and also a removal of a lduw_raw() call from (what is now) hw/display/blizzard_template.h which I suspect is also wrong... Most likely it was an incorrectly squashed patch. Paolo
Re: [Qemu-devel] [PATCH 0/2] Use g_new() friends where that makes obvious sense
Peter Maydell peter.mayd...@linaro.org writes: On 31 January 2014 15:53, Markus Armbruster arm...@redhat.com wrote: PATCH 1/2 fixes a bug found by compiling the original PATCH 2/2. Evidence for my claim that PATCH 2/2 lets the compiler catch more type errors :) Markus Armbruster (2): qga: Fix memory allocation pasto Use g_new() friends where that makes obvious sense 186 files changed, 376 insertions(+), 415 deletions(-) No objection in principle, but I think this is going to be hideous merge pain since it touches a huge range of files. Could you split it up into separate patches that could reasonably go in via the appropriate submaintainer trees? No merge pain at all if you simply rerun the included Coccinelle patch! Splitting the patch may shift some pain from the choke point (you) to submaintainers and me. I don't mind the splitting pain. I do mind the chase the nominal maintainer of obscure corner pain. If you really want it split: what about splitting off just the busy and/or well-maintained subsystems?
[Qemu-devel] qemu-iotest 077 does not terminate with VMDK
Hi, has anyone tried iotest 77 (RMW) with VMDK? for me it does not terminate. Peter
Re: [Qemu-devel] [PATCH 2/2] Use g_new() friends where that makes obvious sense
Eric Blake ebl...@redhat.com writes:
On 01/31/2014 08:53 AM, Markus Armbruster wrote:
g_new(T, n) is safer than g_malloc(sizeof(T) * n) for two reasons.
One, it catches multiplication overflowing size_t. Two, it returns
T * rather than void *, which lets the compiler catch more type
errors.
Patch created with the following Coccinelle patch, with two hunks
dropped from the result:
Looks like a reasonable formula for replacement.
Signed-off-by: Markus Armbruster arm...@redhat.com
---
186 files changed, 375 insertions(+), 414 deletions(-)
Is there any easy way to enforce that this style does not creep back
into the code base? In other words, can you do a followup patch to
checkpatch.pl that flags use of g_malloc[0]?
There are many, many uses of g_malloc(), g_malloc0() and g_realloc()
left. The ones with a size argument of the form sizeof(E) * N could be
rewritten, too. But there are plenty of others.
For the sizeof(E) * N cases, rewriting may not be an improvement. The
pattern p = malloc(sizeof(*p)) is pretty idiomatic.
Having checkpatch.pl match the unwanted patterns sizeof(T), sizeof(T)*N,
and their variations seems infeasible to me, because deciding whether T
is a type is beyond its capabilities.
Ideas?
@@ -658,7 +658,7 @@ int qcow2_snapshot_list(BlockDriverState *bs,
QEMUSnapshotInfo **psn_tab)
return s-nb_snapshots;
}
-sn_tab = g_malloc0(s-nb_snapshots * sizeof(QEMUSnapshotInfo));
+sn_tab = g_new0(QEMUSnapshotInfo, s-nb_snapshots);
for(i = 0; i s-nb_snapshots; i++) {
Should checkpatch.pl care about the spacing after 'for'? Then again,
it's unrelated to your conversion, so omitting a whitespace cleanup from
this patch doesn't hurt.
It already does as far as I can tell, but as Richard said, it only
complains about style violations in changed lines.
+++ b/hw/char/virtio-serial-bus.c
@@ -921,10 +921,8 @@ static void
virtio_serial_device_realize(DeviceState *dev, Error **errp)
QTAILQ_INIT(vser-ports);
vser-bus.max_nr_ports = vser-serial.max_virtserial_ports;
-vser-ivqs = g_malloc(vser-serial.max_virtserial_ports
- * sizeof(VirtQueue *));
-vser-ovqs = g_malloc(vser-serial.max_virtserial_ports
- * sizeof(VirtQueue *));
+vser-ivqs = g_new(VirtQueue *, vser-serial.max_virtserial_ports);
+vser-ovqs = g_new(VirtQueue *, vser-serial.max_virtserial_ports);
I'm impressed at what Coccinelle can rewrite!
Yup. It's quite capable, but its language is weird, and its error
messages can compete with C++ is mysteriousness (thankfully not in
length).
+++ b/ui/keymaps.c
@@ -107,7 +107,7 @@ static kbd_layout_t *parse_keyboard_layout(const
name2keysym_t *table,
}
if (!k)
-k = g_malloc0(sizeof(kbd_layout_t));
+k = g_new0(kbd_layout_t, 1);
for(;;) {
Fixing tab damage while you are at it - nice. And just noticed this is
another instance of no space after 'for', but not worth fixing in this
patch if checkpatch.pl is happy.
I'm trusting coccinelle, and only glanced through random spots of the
patch; but where I looked, I didn't find any problems in the conversion.
I don't know if that means we should add:
Reviewed-by: Eric Blake ebl...@redhat.com
Thank you!
[Qemu-devel] GDB info registers vs. monitor info registers
Hello, I have some trouble to implement a proper system reset for SPARC/LEON. It seems that a Qemu system reset doesn't propagate the state change thoroughly. (gdb) monitor system_reset (gdb) info registers g0 0x0 0 g1 0x40001fa8 1073749928 g2 0x40054f20 1074089760 g3 0x6 6 g4 0x6 6 g5 0x0 0 g6 0x0 0 g7 0x4005d800 1074124800 o0 0x0 0 o1 0x4002c614 1073923604 o2 0x1 1 o3 0x0 0 o4 0x8 8 o5 0x4004c37c 1074054012 sp 0x4005aad8 0x4005aad8 o7 0x4004f6bc 1074067132 l0 0xf3c3 -218103613 l1 0x40013490 1073820816 l2 0x40013494 1073820820 l3 0x80 128 l4 0xf3c3 -218103613 l5 0xf3c3 -218103613 l6 0x3e 62 l7 0x3f 63 i0 0x40054f20 1074089760 i1 0x43fffda8 1140850088 i2 0x1 1 i3 0x0 0 i4 0x55493120 1430860064 i5 0x100256 fp 0x4005abc0 0x4005abc0 i7 0x4003a11c 1073979676 y 0x0 0 psr0xf34000e3 [ #0 #1 ET PS S #22 #24 #25 #28 #29 #30 #31 ] wim0x20 32 tbr0x4060 1073741920 pc 0x40001fb0 0x40001fb0 Init(rtems_task_argument)+8 npc0x40001fb4 0x40001fb4 Init(rtems_task_argument)+12 fsr0x0 [ ] csr0x0 0 (gdb) monitor info registers pc: 4000 npc: 4004 %g0-7: %o0-7: 0001 0002 0003 0004 0005 0006 0007 %l0-7: 0008 0009 000a 000b 000c 000d 000e 000f %i0-7: 0010 0011 0012 0013 0014 0015 0016 0017 %f00: %f08: %f16: %f24: psr: f3c0 (icc: SPE: SP-) wim: 0001 fsr: y: A monitor system_reset triggers a call to main_cpu_reset (hw/sparc/leon3.c:51) and the CPUSPARCState is set up correctly, but via GDB I see no state change and the program executes as if nothing happened. Does this work as intended? -- Sebastian Huber, embedded brains GmbH Address : Dornierstr. 4, D-82178 Puchheim, Germany Phone : +49 89 189 47 41-16 Fax : +49 89 189 47 41-09 E-Mail : sebastian.hu...@embedded-brains.de PGP : Public key available on request. Diese Nachricht ist keine geschäftliche Mitteilung im Sinne des EHUG.
Re: [Qemu-devel] [PATCH 0/2] qtest: don't leak pid files and UNIX domain sockets
Hi, ERROR:/home/petmay01/linaro/qemu-for-merges/tests/libqtest.c:71:init_socket: assertion failed (ret != -1): (-1 != -1) FAIL GTester: last random seed: R02S79ea313790bc9a8b21d9af5ed55c2fff (pid=5080) /i386/qom/pc:OK /i386/qom/isapc: OK /i386/qom/q35: OK FAIL: tests/qom-test but this patch series doesn't actually say that's what it's for, so does it fix that kind of error? I might help but I'm not sure. The assertion failure you posted happened on Gerd's buildslaves, I haven't reproduced it myself and don't really understand what's wrong there. It seems to have stopped happening on the buildslaves, not sure whenever this is just luck or whenever it was fixed by some patch. However, I noticed that my /tmp is getting cluttered with pid files and UNIX domain sockets from qtest. I also noticed that QEMU processes are left running (this is not fixed by my series) when tests fail. So this series at least stops us from cluttering /tmp. Removing the clutter from /tmp unbroke the buildslaves, so not cluttering tmp certainly is a good thing. cheers, Gerd
Re: [Qemu-devel] [PATCH 0/2] Use g_new() friends where that makes obvious sense
On 3 February 2014 08:40, Markus Armbruster arm...@redhat.com wrote: Peter Maydell peter.mayd...@linaro.org writes: On 31 January 2014 15:53, Markus Armbruster arm...@redhat.com wrote: 186 files changed, 376 insertions(+), 415 deletions(-) No objection in principle, but I think this is going to be hideous merge pain since it touches a huge range of files. Could you split it up into separate patches that could reasonably go in via the appropriate submaintainer trees? No merge pain at all if you simply rerun the included Coccinelle patch! Yes, but that requires me to find, install, learn about and use Coccinelle. Splitting the patch may shift some pain from the choke point (you) to submaintainers and me. I don't mind the splitting pain. I do mind the chase the nominal maintainer of obscure corner pain. If you really want it split: what about splitting off just the busy and/or well-maintained subsystems? Yes, that's fine. You can put the miscellaneous leftovers parts through trivial if you like. I just dislike single touches-entire-world patches if they're not absolutely necessary. thanks -- PMM
[Qemu-devel] [PATCH] sparc/leon3: Initialize stack pointer
A lot of real world LEON3 systems are shipped with the GRMON boot
loader. This boot loader initializes the stack pointer with the end of
RAM address. The application can use this to detect the RAM size of a
particular board variant.
Signed-off-by: Sebastian Huber sebastian.hu...@embedded-brains.de
---
hw/sparc/leon3.c |3 +++
1 files changed, 3 insertions(+), 0 deletions(-)
diff --git a/hw/sparc/leon3.c b/hw/sparc/leon3.c
index c583c3d..c16e9e4 100644
--- a/hw/sparc/leon3.c
+++ b/hw/sparc/leon3.c
@@ -45,6 +45,7 @@
typedef struct ResetData {
SPARCCPU *cpu;
uint32_t entry;/* save kernel entry in case of reset */
+target_ulong sp;/* initial stack pointer */
} ResetData;
static void main_cpu_reset(void *opaque)
@@ -58,6 +59,7 @@ static void main_cpu_reset(void *opaque)
cpu-halted = 0;
env-pc = s-entry;
env-npc= s-entry + 4;
+env-regbase[6] = s-sp;
}
void leon3_irq_ack(void *irq_manager, int intno)
@@ -133,6 +135,7 @@ static void leon3_generic_hw_init(QEMUMachineInitArgs *args)
/* Reset data */
reset_info= g_malloc0(sizeof(ResetData));
reset_info-cpu = cpu;
+reset_info-sp= 0x4000 + ram_size;
qemu_register_reset(main_cpu_reset, reset_info);
/* Allocate IRQ manager */
--
1.7.7
[Qemu-devel] [PATCHv9 2/5] qemu-iotests: change _supported_proto to file for various tests
all these tests do anything of the following and thus fail with any protocol other than file: - the tests use rm, cp or mv shell commands which only work on file - the tests use qcow2.py - the images construct new filenames (e.g. backing file names) and the logic is broken for anything else than file Signed-off-by: Peter Lieven p...@kamp.de --- tests/qemu-iotests/013 |2 +- tests/qemu-iotests/014 |2 +- tests/qemu-iotests/018 |2 +- tests/qemu-iotests/019 |2 +- tests/qemu-iotests/023 |2 +- tests/qemu-iotests/024 |2 +- tests/qemu-iotests/026 |2 +- tests/qemu-iotests/028 |2 +- tests/qemu-iotests/031 |2 +- tests/qemu-iotests/034 |2 +- tests/qemu-iotests/036 |2 +- tests/qemu-iotests/037 |2 +- tests/qemu-iotests/038 |2 +- tests/qemu-iotests/039 |2 +- tests/qemu-iotests/043 |2 +- tests/qemu-iotests/046 |2 +- tests/qemu-iotests/052 |2 +- tests/qemu-iotests/054 |2 +- tests/qemu-iotests/059 |2 +- tests/qemu-iotests/060 |2 +- tests/qemu-iotests/061 |2 +- tests/qemu-iotests/063 |2 +- tests/qemu-iotests/069 |2 +- tests/qemu-iotests/071 |2 +- tests/qemu-iotests/072 |2 +- 25 files changed, 25 insertions(+), 25 deletions(-) diff --git a/tests/qemu-iotests/013 b/tests/qemu-iotests/013 index 389f4b8..ea3cab9 100755 --- a/tests/qemu-iotests/013 +++ b/tests/qemu-iotests/013 @@ -41,7 +41,7 @@ trap _cleanup; exit \$status 0 1 2 3 15 # much of this could be generic for any format supporting compression. _supported_fmt qcow qcow2 -_supported_proto generic +_supported_proto file _supported_os Linux TEST_OFFSETS=0 4294967296 diff --git a/tests/qemu-iotests/014 b/tests/qemu-iotests/014 index 0edeb4b..b23c2db 100755 --- a/tests/qemu-iotests/014 +++ b/tests/qemu-iotests/014 @@ -43,7 +43,7 @@ trap _cleanup; exit \$status 0 1 2 3 15 # much of this could be generic for any format supporting snapshots _supported_fmt qcow2 -_supported_proto generic +_supported_proto file _supported_os Linux TEST_OFFSETS=0 4294967296 diff --git a/tests/qemu-iotests/018 b/tests/qemu-iotests/018 index 6f7f054..d8a7d43 100755 --- a/tests/qemu-iotests/018 +++ b/tests/qemu-iotests/018 @@ -41,7 +41,7 @@ trap _cleanup; exit \$status 0 1 2 3 15 # Any format supporting backing files _supported_fmt qcow qcow2 vmdk qed -_supported_proto generic +_supported_proto file _supported_os Linux _unsupported_imgopts subformat=monolithicFlat subformat=twoGbMaxExtentFlat diff --git a/tests/qemu-iotests/019 b/tests/qemu-iotests/019 index b43e70f..e67445c 100755 --- a/tests/qemu-iotests/019 +++ b/tests/qemu-iotests/019 @@ -45,7 +45,7 @@ trap _cleanup; exit \$status 0 1 2 3 15 # Any format supporting backing files _supported_fmt qcow qcow2 vmdk qed -_supported_proto generic +_supported_proto file _supported_os Linux _unsupported_imgopts subformat=monolithicFlat \ subformat=twoGbMaxExtentFlat \ diff --git a/tests/qemu-iotests/023 b/tests/qemu-iotests/023 index 090ed23..9ad06b9 100755 --- a/tests/qemu-iotests/023 +++ b/tests/qemu-iotests/023 @@ -41,7 +41,7 @@ trap _cleanup; exit \$status 0 1 2 3 15 # much of this could be generic for any format supporting compression. _supported_fmt qcow qcow2 -_supported_proto generic +_supported_proto file _supported_os Linux TEST_OFFSETS=0 4294967296 diff --git a/tests/qemu-iotests/024 b/tests/qemu-iotests/024 index be974f0..9bf99e1 100755 --- a/tests/qemu-iotests/024 +++ b/tests/qemu-iotests/024 @@ -43,7 +43,7 @@ trap _cleanup; exit \$status 0 1 2 3 15 # Currently only qcow2 and qed support rebasing _supported_fmt qcow2 qed -_supported_proto generic +_supported_proto file _supported_os Linux CLUSTER_SIZE=65536 diff --git a/tests/qemu-iotests/026 b/tests/qemu-iotests/026 index c9c5f83..df2884b 100755 --- a/tests/qemu-iotests/026 +++ b/tests/qemu-iotests/026 @@ -42,7 +42,7 @@ trap _cleanup; exit \$status 0 1 2 3 15 # Currently only qcow2 supports rebasing _supported_fmt qcow2 -_supported_proto generic +_supported_proto file _supported_os Linux _default_cache_mode writethrough _supported_cache_modes writethrough none diff --git a/tests/qemu-iotests/028 b/tests/qemu-iotests/028 index 93a9fa6..a99e4fa 100755 --- a/tests/qemu-iotests/028 +++ b/tests/qemu-iotests/028 @@ -45,7 +45,7 @@ trap _cleanup; exit \$status 0 1 2 3 15 # Any format supporting backing files except vmdk and qcow which do not support # smaller backing files. _supported_fmt qcow2 qed -_supported_proto generic +_supported_proto file _supported_os Linux # Choose a size that is not necessarily a cluster size multiple for image diff --git a/tests/qemu-iotests/031 b/tests/qemu-iotests/031 index c9070b0..1d920ea 100755 --- a/tests/qemu-iotests/031 +++ b/tests/qemu-iotests/031 @@ -41,7 +41,7 @@ trap _cleanup; exit \$status 0 1 2 3 15 # This tests qcow2-specific low-level functionality _supported_fmt qcow2 -_supported_proto generic +_supported_proto file _supported_os Linux
[Qemu-devel] [PATCHv9 1/5] block: add native support for NFS
This patch adds native support for accessing images on NFS
shares without the requirement to actually mount the entire
NFS share on the host.
NFS Images can simply be specified by an url of the form:
nfs://host/export/filename[?param=value[param2=value2[...]]]
For example:
qemu-img create -f qcow2 nfs://10.0.0.1/qemu-images/test.qcow2
You need LibNFS from Ronnie Sahlberg available at:
git://github.com/sahlberg/libnfs.git
for this to work.
During configure it is automatically probed for libnfs and support
is enabled on-the-fly. You can forbid or enforce libnfs support
with --disable-libnfs or --enable-libnfs respectively.
Due to NFS restrictions you might need to execute your binaries
as root, allow them to open priviledged ports (1024) or specify
insecure option on the NFS server.
For additional information on ROOT vs. non-ROOT operation and URL
format + parameters see:
https://raw.github.com/sahlberg/libnfs/master/README
Supported by qemu are the uid, gid and tcp-syncnt URL parameters.
LibNFS currently support NFS version 3 only.
Signed-off-by: Peter Lieven p...@kamp.de
---
MAINTAINERS |5 +
block/Makefile.objs |1 +
block/nfs.c | 439 +++
configure | 26 +++
qapi-schema.json|1 +
5 files changed, 472 insertions(+)
create mode 100644 block/nfs.c
diff --git a/MAINTAINERS b/MAINTAINERS
index fb53242..f8411f9 100644
--- a/MAINTAINERS
+++ b/MAINTAINERS
@@ -936,6 +936,11 @@ M: Peter Lieven p...@kamp.de
S: Supported
F: block/iscsi.c
+NFS
+M: Peter Lieven p...@kamp.de
+S: Maintained
+F: block/nfs.c
+
SSH
M: Richard W.M. Jones rjo...@redhat.com
S: Supported
diff --git a/block/Makefile.objs b/block/Makefile.objs
index 4e8c91e..e254a21 100644
--- a/block/Makefile.objs
+++ b/block/Makefile.objs
@@ -12,6 +12,7 @@ block-obj-$(CONFIG_LINUX_AIO) += linux-aio.o
ifeq ($(CONFIG_POSIX),y)
block-obj-y += nbd.o nbd-client.o sheepdog.o
block-obj-$(CONFIG_LIBISCSI) += iscsi.o
+block-obj-$(CONFIG_LIBNFS) += nfs.o
block-obj-$(CONFIG_CURL) += curl.o
block-obj-$(CONFIG_RBD) += rbd.o
block-obj-$(CONFIG_GLUSTERFS) += gluster.o
diff --git a/block/nfs.c b/block/nfs.c
new file mode 100644
index 000..ef731f0
--- /dev/null
+++ b/block/nfs.c
@@ -0,0 +1,439 @@
+/*
+ * QEMU Block driver for native access to files on NFS shares
+ *
+ * Copyright (c) 2014 Peter Lieven p...@kamp.de
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a copy
+ * of this software and associated documentation files (the Software), to
deal
+ * in the Software without restriction, including without limitation the rights
+ * to use, copy, modify, merge, publish, distribute, sublicense, and/or sell
+ * copies of the Software, and to permit persons to whom the Software is
+ * furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in
+ * all copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED AS IS, WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL
+ * THE AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
+ * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING
FROM,
+ * OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN
+ * THE SOFTWARE.
+ */
+
+#include config-host.h
+
+#include poll.h
+#include qemu-common.h
+#include qemu/config-file.h
+#include qemu/error-report.h
+#include block/block_int.h
+#include trace.h
+#include qemu/iov.h
+#include qemu/uri.h
+#include sysemu/sysemu.h
+#include nfsc/libnfs.h
+
+typedef struct NFSClient {
+struct nfs_context *context;
+struct nfsfh *fh;
+int events;
+bool has_zero_init;
+} NFSClient;
+
+typedef struct NFSRPC {
+int ret;
+int complete;
+QEMUIOVector *iov;
+struct stat *st;
+Coroutine *co;
+QEMUBH *bh;
+} NFSRPC;
+
+static void nfs_process_read(void *arg);
+static void nfs_process_write(void *arg);
+
+static void nfs_set_events(NFSClient *client)
+{
+int ev = nfs_which_events(client-context);
+if (ev != client-events) {
+qemu_aio_set_fd_handler(nfs_get_fd(client-context),
+ (ev POLLIN) ? nfs_process_read : NULL,
+ (ev POLLOUT) ? nfs_process_write : NULL,
+ client);
+
+}
+client-events = ev;
+}
+
+static void nfs_process_read(void *arg)
+{
+NFSClient *client = arg;
+nfs_service(client-context, POLLIN);
+nfs_set_events(client);
+}
+
+static void nfs_process_write(void *arg)
+{
+NFSClient *client = arg;
+nfs_service(client-context, POLLOUT);
+nfs_set_events(client);
+}
+
+static void nfs_co_init_task(NFSClient *client, NFSRPC *task)
+{
+*task = (NFSRPC) {
+.co = qemu_coroutine_self(),
+};
+}
[Qemu-devel] [PATCHv9 0/5] block: add native support for NFS
This adds v9 of the NFS protocol driver + qemu-iotest adjustments. v8-v9: - change _supported_proto for qemu-iotests 071 072 from generic to file v7-v8: - bump libnfs requirement to 1.9.2 - renamed NFSRPC-status to NFSRPC-ret [Benoit] - avoid double invocation of nfs_client_close in nfs_file_create [Benoit] - fixed error propagation in nfs_file_open [Benoit] v6-v7: - removed unsused and partially non-existing libnfs header file includes [Stefan] - adjusted usage of qemu_opts_create_nofail [Stefan] - removed patch to fix output of qemu-iotest 067 - only blacklist qemu-iotest 020 for NFS do not change proto from generic to file [Fam] v5-v6: - use internal qemu function to parse the NFS url [Kevin] - zero pad short reads [Kevin, Paolo] - added qemu-iotests patches for basic nfs protocol support v4-v5: - disussed with Ronnie and decided to move URL + Paramter parsing to LibNFS. This allows for URL parameter processing directly in LibNFS without altering the qemu NFS block driver. This bumps the version requirement for LibNFS to 1.9.0 though. - added a pointer to the LibNFS readme where additional information about ROOT privilidge requirements can be found as this raised a few concerns. - removed a trailing dot in an error statement [Fam]. v3-v4: - finally added full implementation of bdrv_get_allocated_file_size [Stefan] - removed trailing \n from error statements [Stefan] v2-v3: - rebased the stefanha/block - use pkg_config to check for libnfs (ignoring cflags which are broken in 1.8.0) [Stefan] - fixed NFSClient declaration [Stefan] - renamed Task variables to task [Stefan] - renamed NFSTask to NFSRPC [Ronnie] - do not update bs-total_sectors in nfs_co_writev [Stefan] - return -ENOMEM on all async call failures [Stefan,Ronnie] - fully implement ftruncate - use util/uri.c for URL parsing [Stefan] - reworked nfs_file_open_common to nfs_client_open which works on NFSClient [Stefan] - added a comment ot the connect message that libnfs support NFSv3 only at the moment. - DID NOT add full implementation of bdrv_get_allocated_file_size because we are not in a coroutine context and I cannot do an async call here. I could do a sync call if there would be a guarantee that no requests are in flight. [Stefan] v1-v2: - fixed block/Makefile.objs [Ronnie] - do not always register a read handler [Ronnie] - add support for reading beyond EOF [Fam] - fixed struct and paramter naming [Fam] - fixed overlong lines and whitespace errors [Fam] - return return status from libnfs whereever possible [Fam] - added comment why we set allocated_file_size to -ENOTSUP after write [Fam] - avoid segfault when parsing filname [Fam] - remove unused close_bh from NFSClient [Fam] - avoid dividing and mutliplying total_size by BDRV_SECTOR_SIZE in nfs_file_create [Fam] Peter Lieven (5): block: add native support for NFS qemu-iotests: change _supported_proto to file for various tests qemu-iotests: blacklist test 020 for NFS protocol qemu-iotests: enable test 016 and 025 to work with NFS protocol qemu-iotests: enable support for NFS protocol MAINTAINERS |5 + block/Makefile.objs |1 + block/nfs.c | 439 ++ configure| 26 +++ qapi-schema.json |1 + tests/qemu-iotests/013 |2 +- tests/qemu-iotests/014 |2 +- tests/qemu-iotests/016 |2 +- tests/qemu-iotests/018 |2 +- tests/qemu-iotests/019 |2 +- tests/qemu-iotests/020 |5 + tests/qemu-iotests/023 |2 +- tests/qemu-iotests/024 |2 +- tests/qemu-iotests/025 |2 +- tests/qemu-iotests/026 |2 +- tests/qemu-iotests/028 |2 +- tests/qemu-iotests/031 |2 +- tests/qemu-iotests/034 |2 +- tests/qemu-iotests/036 |2 +- tests/qemu-iotests/037 |2 +- tests/qemu-iotests/038 |2 +- tests/qemu-iotests/039 |2 +- tests/qemu-iotests/043 |2 +- tests/qemu-iotests/046 |2 +- tests/qemu-iotests/052 |2 +- tests/qemu-iotests/054 |2 +- tests/qemu-iotests/059 |2 +- tests/qemu-iotests/060 |2 +- tests/qemu-iotests/061 |2 +- tests/qemu-iotests/063 |2 +- tests/qemu-iotests/069 |2 +- tests/qemu-iotests/071 |2 +- tests/qemu-iotests/072 |2 +- tests/qemu-iotests/common| 22 ++- tests/qemu-iotests/common.rc |3 + 35 files changed, 526 insertions(+), 30 deletions(-) create mode 100644 block/nfs.c -- 1.7.9.5
[Qemu-devel] [PATCHv9 4/5] qemu-iotests: enable test 016 and 025 to work with NFS protocol
Signed-off-by: Peter Lieven p...@kamp.de --- tests/qemu-iotests/016 |2 +- tests/qemu-iotests/025 |2 +- 2 files changed, 2 insertions(+), 2 deletions(-) diff --git a/tests/qemu-iotests/016 b/tests/qemu-iotests/016 index b87a32b..7ea9e94 100755 --- a/tests/qemu-iotests/016 +++ b/tests/qemu-iotests/016 @@ -39,7 +39,7 @@ trap _cleanup; exit \$status 0 1 2 3 15 . ./common.filter _supported_fmt raw -_supported_proto file sheepdog +_supported_proto file sheepdog nfs _supported_os Linux diff --git a/tests/qemu-iotests/025 b/tests/qemu-iotests/025 index 9426c93..a5f45b4 100755 --- a/tests/qemu-iotests/025 +++ b/tests/qemu-iotests/025 @@ -40,7 +40,7 @@ trap _cleanup; exit \$status 0 1 2 3 15 . ./common.pattern _supported_fmt raw qcow2 qed -_supported_proto file sheepdog rbd +_supported_proto file sheepdog rbd nfs _supported_os Linux echo === Creating image -- 1.7.9.5
Re: [Qemu-devel] [PATCHv4 0/6] ui/vnc: update optimizations
On 08.01.2014 10:08, Peter Lieven wrote: this series includes several optimizations for the ui/vnc guest to server and server to client update cycles. comments/reviews appreciated. v3-v4: - patch 1,4,6 unchanged - patch 2: fix 2 missing changes from 16 - VNC_DIRTY_PIXELS_PER_BIT [Wenchao] - patch 3: fixed missing roundup of width in vnc_dpy_switch [Wenchao] - patch 3: fixed x offset calculation in vnc_refresh_server_surface - patch 5: fixed missing roundup of width in vnc_dpy_update, cleaned up function v2-v3: - fixed checkpatch warnings [Stefan] - fixed an old coding style violation [Stefan] v1-v2: - new patches 2,4,5 - patch3: added performance test [Anthony] - patch3: further optimized the vnc_update_client by searching for the next zero bit with find_next_zero_bit. - patch3: further optimized vnc_dpy_switch by using bitmap_set to mask bits dirty. Peter Lieven (6): ui/vnc: introduce VNC_DIRTY_PIXELS_PER_BIT macro ui/vnc: derive cmp_bytes from VNC_DIRTY_PIXELS_PER_BIT ui/vnc: optimize dirty bitmap tracking ui/vnc: optimize clearing in find_and_clear_dirty_height() ui/vnc: optimize setting in vnc_dpy_update() ui/vnc: disable adaptive update calculations if not needed ui/vnc.c | 206 +++--- ui/vnc.h | 10 ++- 2 files changed, 127 insertions(+), 89 deletions(-) PingV2 ;-) Peter
[Qemu-devel] [PATCHv9 5/5] qemu-iotests: enable support for NFS protocol
Signed-off-by: Peter Lieven p...@kamp.de --- tests/qemu-iotests/common| 22 +++--- tests/qemu-iotests/common.rc |3 +++ 2 files changed, 22 insertions(+), 3 deletions(-) diff --git a/tests/qemu-iotests/common b/tests/qemu-iotests/common index 8b4e22c..5795358 100644 --- a/tests/qemu-iotests/common +++ b/tests/qemu-iotests/common @@ -144,10 +144,12 @@ check options -vpctest vpc -vhdx test vhdx -vmdk test vmdk +-file test file (default) -rbdtest rbd -sheepdog test sheepdog -nbdtest nbd -sshtest ssh +-nfstest nfs -xdiff graphical mode diff -nocacheuse O_DIRECT on backing file -misalign misalign memory allocations @@ -211,22 +213,36 @@ testlist options xpand=false ;; +-file) +IMGPROTO=file +xpand=false +;; + -rbd) IMGPROTO=rbd xpand=false ;; + -sheepdog) IMGPROTO=sheepdog xpand=false ;; + -nbd) IMGPROTO=nbd xpand=false ;; + -ssh) IMGPROTO=ssh xpand=false ;; + +-nfs) +IMGPROTO=nfs +xpand=false +;; + -nocache) CACHEMODE=none CACHEMODE_IS_DEFAULT=false @@ -238,10 +254,10 @@ testlist options xpand=false ;; --valgrind) -valgrind=true +-valgrind) +valgrind=true xpand=false -;; +;; -g)# -g group ... pick from group file group=true diff --git a/tests/qemu-iotests/common.rc b/tests/qemu-iotests/common.rc index 0f68156..71e9a74 100644 --- a/tests/qemu-iotests/common.rc +++ b/tests/qemu-iotests/common.rc @@ -61,6 +61,9 @@ elif [ $IMGPROTO = nbd ]; then elif [ $IMGPROTO = ssh ]; then TEST_IMG_FILE=$TEST_DIR/t.$IMGFMT TEST_IMG=ssh://127.0.0.1$TEST_IMG_FILE +elif [ $IMGPROTO = nfs ]; then +TEST_DIR=nfs://127.0.0.1/$TEST_DIR +TEST_IMG=$TEST_DIR/t.$IMGFMT else TEST_IMG=$IMGPROTO:$TEST_DIR/t.$IMGFMT fi -- 1.7.9.5
[Qemu-devel] [PATCHv9 3/5] qemu-iotests: blacklist test 020 for NFS protocol
reopening is currently not supported. Signed-off-by: Peter Lieven p...@kamp.de --- tests/qemu-iotests/020 |5 + 1 file changed, 5 insertions(+) diff --git a/tests/qemu-iotests/020 b/tests/qemu-iotests/020 index 73a0429..2f258dc 100755 --- a/tests/qemu-iotests/020 +++ b/tests/qemu-iotests/020 @@ -49,6 +49,11 @@ _unsupported_imgopts subformat=monolithicFlat \ subformat=twoGbMaxExtentFlat \ subformat=twoGbMaxExtentSparse +# NFS does not support bdrv_reopen_prepare thus qemu-img commit fails. +if [ $IMGPROTO = nfs ]; then +_notrun image protocol $IMGPROTO does not support bdrv_commit +fi + TEST_OFFSETS=0 4294967296 _make_test_img 6G -- 1.7.9.5
Re: [Qemu-devel] 9pfs troubles (was Re: [PATCH 1/4] hw/9pfs: fix error handing in local_ioc_getversion())
Michael S. Tsirkin m...@redhat.com writes: Haven't used 9pfs in a while. I thought these patches are a good time to play with it some more. I have encountered two issues. What I'm doing: host: qemu a75143eda2ddf581b51e96c000974bcdfe2cbd10. /scm/qemu/x86_64-softmmu/qemu-system-x86_64 -enable-kvm -m 1g -cpu kvm64 -smp 2 f20-x64.qcow2 -netdev user,id=foo -redir tcp:8022::22 -device virtio-net,netdev=foo -serial stdio -fsdev local,security_model=none,id=fsdev0,path=/lib/modules/ -device virtio-9p-pci,id=fs0,fsdev=fsdev0,mount_tag=libmodulesshare -fsdev local,security_model=none,id=fsdev1,path=/boot -device virtio-9p-pci,id=fs1,fsdev=fsdev0,mount_tag=bootshare -no-reboot -snapshot guest: Fedora 20 added this in /etc/fstab: bootshare /share/boot 9p trans=virtio,version=9p2000.L 0 0 libmodulesshare /share/lib/modules 9p trans=virtio,version=9p2000.L 0 0 I have encountered two issues: 1. mount failure on boot If I try to mount on boot through fstab, I get: [2.270157] 9pnet: Could not find request transport: virtio [2.270158] 9pnet: Could not find request transport: virtio Missing 9pnet_virtio.ko module ? If I then re-try mount, it succeeds immediately! Some kind of dependency issue? 2. files immediately in the mounted directory aren't visible on the guest under /share/boot. For example, files under /boot on host are not visible on guest, files under child directories seem visible. can you share more details on this ? /boot permissions. ls -al output on host etc. -aneesh
Re: [Qemu-devel] [PATCH 20/42] input: mouse: add graphic_rotate support
Hi,
+switch (graphic_rotate) {
+case 90:
+if (evt-abs-axis == INPUT_AXIS_X) {
+evt-abs-axis = INPUT_AXIS_Y;
+}
+if (evt-abs-axis == INPUT_AXIS_Y) {
Need else here, same for case 270.
Why?
+evt-abs-axis = INPUT_AXIS_X;
+evt-abs-axis = INPUT_EVENT_ABS_SIZE - 1 - evt-abs-axis;
-value here, not -axis.
Oops, indeed.
It looks like doing it right for relative is easy:
But what is the point when this isn't used anyway?
if (graphic_rotate == 0) {
return;
}
if (move-axis != INPUT_AXIS_X move-axis != INPUT_AXIS_Y) {
return;
}
if ((graphic_rotate = 180 move-axis == INPUT_AXIS_X) ||
(graphic_rotate = 180 move-axis == INPUT_AXIS_Y)) {
if (kind == INPUT_EVENT_KIND_ABS) {
move-value = INPUT_EVENT_ABS_SIZE - 1 - move-value;
} else {
move-value = -move-value;
}
}
if (graphic_rotate == 90 || graphic_rotate == 270) {
move-axis ^= INPUT_AXIS_X ^ INPUT_AXIS_Y;
}
Saves a few lines but it's a bit harder to figure what is going on ...
cheers,
Gerd
Re: [Qemu-devel] [PATCH 06/42] input: qapi: define event types
Hi, s/alot/a lot/ s/splitted/split/ Applied. +## +# @InputBtnEvent Should we spell this out? I'm okay with the abbreviation if it makes it easier to fit 80 columns, but Button is nicer to read than Btn. I'm sort-of following linux input layer convention here, which uses btn for the buttons too. cheers, Gerd
Re: [Qemu-devel] [PATCH V6 8/8] block: Use graph node name as reference in bdrv_file_open().
Am 31.01.2014 um 22:37 hat Benoît Canet geschrieben: Le Friday 31 Jan 2014 à 21:32:34 (+0100), Max Reitz a écrit : On 28.01.2014 01:04, Benoît Canet wrote: Le Monday 27 Jan 2014 à 20:11:59 (+0100), Max Reitz a écrit : On 27.01.2014 15:36, Benoît Canet wrote: Le Friday 24 Jan 2014 à 15:54:39 (+0100), Max Reitz a écrit : On 24.01.2014 15:48, Kevin Wolf wrote: Am 24.01.2014 um 14:37 hat Max Reitz geschrieben: On 24.01.2014 14:26, Kevin Wolf wrote: Am 23.01.2014 um 21:31 hat Benoît Canet geschrieben: Signed-off-by: Benoit Canet ben...@irqsave.net --- block.c | 6 +++--- 1 file changed, 3 insertions(+), 3 deletions(-) I'm not going to merge this one yet. It breaks qemu-iotests case 071, which would have to be adapted. However, first of all I'd like to hear the opinions of at least Eric and Max on what BlockRef should really refer to. I think node names make most sense, but perhaps it's a bit inconvenient and the command line should default to node-name = id when id is set, but node-name isn't? The QAPI schema is pretty clear about this: “references the ID of an existing block device.” Sure, that's because I wrote that text before we had a node name. However, in 1.7 references didn't work yet, so we still have all freedom to change the interface as we like. Yes, that's right. However, if the ID cannot be found, I think we should interpret it as a reference to the node name. Therefore, I'd first try bdrv_find() and if that returns NULL, try again with bdrv_find_node(). I think I would prefer to avoid such ambiguities. Otherwise a management tool that wants to use the node name needs to check first if it's not already used as a device name somewhere else and would therefore operate on the wrong device. On the other hand, a management tool using the same names for devices and nodes just gets what it deserves. Perhaps we should use a common namespace for both, i.e. you get an error if you try to assign a node name that is already a device name and vice versa? This is what I would go for. However, then I don't really know why we should separate the ID and the node name in the first place (although that's probably because I haven't followed the discussion around node names). Max Ping, I still want to make quorum merge. What should be done for the references ? Best regards Benoît My only problem is that I don't really know what IDs are for, then. ;-) From the understanding I have ID are for block backend top level bds and node-name naming all the bds burried in the graph. So my personal opinion would be to relax the constraint on bdrv_lookup_bs and use it for references. Kevin Max: what do you think of this scheme ? I agree. For example, we could change the constraint to report an error only if both ID and node name are actually valid (and point to different devices), that is, bdrv_find() and bdrv_find_node() return different non-NULL values. Ok I will write patch doing this on top of quorum patches. Yes, I think allowing bdrv_lookup_bs() to find both node names and device names makes sense. I would still use a common namespace and forbid using the same name for a device and a node. Kevin
[Qemu-devel] [PATCH v4 01/22] exec: Make tb_invalidate_phys_addr input an AS
From: Edgar E. Iglesias edgar.igles...@xilinx.com
No functional change.
Signed-off-by: Edgar E. Iglesias edgar.igles...@xilinx.com
---
exec.c| 3 ++-
include/exec/exec-all.h | 2 +-
target-xtensa/op_helper.c | 3 ++-
translate-all.c | 4 ++--
4 files changed, 7 insertions(+), 5 deletions(-)
diff --git a/exec.c b/exec.c
index 9ad0a4b..fc8ae65 100644
--- a/exec.c
+++ b/exec.c
@@ -488,7 +488,8 @@ static void breakpoint_invalidate(CPUState *cpu,
target_ulong pc)
{
hwaddr phys = cpu_get_phys_page_debug(cpu, pc);
if (phys != -1) {
-tb_invalidate_phys_addr(phys | (pc ~TARGET_PAGE_MASK));
+tb_invalidate_phys_addr(address_space_memory,
+phys | (pc ~TARGET_PAGE_MASK));
}
}
#endif
diff --git a/include/exec/exec-all.h b/include/exec/exec-all.h
index 3b03cbf..9d6c36b 100644
--- a/include/exec/exec-all.h
+++ b/include/exec/exec-all.h
@@ -102,7 +102,7 @@ void tlb_flush(CPUArchState *env, int flush_global);
void tlb_set_page(CPUArchState *env, target_ulong vaddr,
hwaddr paddr, int prot,
int mmu_idx, target_ulong size);
-void tb_invalidate_phys_addr(hwaddr addr);
+void tb_invalidate_phys_addr(AddressSpace *as, hwaddr addr);
#else
static inline void tlb_flush_page(CPUArchState *env, target_ulong addr)
{
diff --git a/target-xtensa/op_helper.c b/target-xtensa/op_helper.c
index cf97025..89a72b5 100644
--- a/target-xtensa/op_helper.c
+++ b/target-xtensa/op_helper.c
@@ -29,6 +29,7 @@
#include helper.h
#include qemu/host-utils.h
#include exec/softmmu_exec.h
+#include exec/address-spaces.h
static void do_unaligned_access(CPUXtensaState *env,
target_ulong addr, int is_write, int is_user, uintptr_t retaddr);
@@ -90,7 +91,7 @@ static void tb_invalidate_virtual_addr(CPUXtensaState *env,
uint32_t vaddr)
int ret = xtensa_get_physical_addr(env, false, vaddr, 2, 0,
paddr, page_size, access);
if (ret == 0) {
-tb_invalidate_phys_addr(paddr);
+tb_invalidate_phys_addr(address_space_memory, paddr);
}
}
diff --git a/translate-all.c b/translate-all.c
index 543e1ff..1ac0246 100644
--- a/translate-all.c
+++ b/translate-all.c
@@ -1357,13 +1357,13 @@ static TranslationBlock *tb_find_pc(uintptr_t tc_ptr)
}
#if defined(TARGET_HAS_ICE) !defined(CONFIG_USER_ONLY)
-void tb_invalidate_phys_addr(hwaddr addr)
+void tb_invalidate_phys_addr(AddressSpace *as, hwaddr addr)
{
ram_addr_t ram_addr;
MemoryRegion *mr;
hwaddr l = 1;
-mr = address_space_translate(address_space_memory, addr, addr, l,
false);
+mr = address_space_translate(as, addr, addr, l, false);
if (!(memory_region_is_ram(mr)
|| memory_region_is_romd(mr))) {
return;
--
1.8.1.2
[Qemu-devel] [PATCH v4 03/22] exec: Always initialize MemorySection address spaces
From: Edgar E. Iglesias edgar.igles...@xilinx.com
Signed-off-by: Edgar E. Iglesias edgar.igles...@xilinx.com
---
exec.c | 2 ++
1 file changed, 2 insertions(+)
diff --git a/exec.c b/exec.c
index 8a90867..3814a7e 100644
--- a/exec.c
+++ b/exec.c
@@ -877,6 +877,7 @@ static void register_subpage(AddressSpaceDispatch *d,
MemoryRegionSection *secti
if (!(existing-mr-subpage)) {
subpage = subpage_init(d-as, base);
+subsection.address_space = d-as;
subsection.mr = subpage-iomem;
phys_page_set(d, base TARGET_PAGE_BITS, 1,
phys_section_add(d-map, subsection));
@@ -1722,6 +1723,7 @@ static subpage_t *subpage_init(AddressSpace *as, hwaddr
base)
static uint16_t dummy_section(PhysPageMap *map, MemoryRegion *mr)
{
MemoryRegionSection section = {
+.address_space = address_space_memory,
.mr = mr,
.offset_within_address_space = 0,
.offset_within_region = 0,
--
1.8.1.2
[Qemu-devel] [PATCH v4 02/22] exec: Make iotlb_to_region input an AS
From: Edgar E. Iglesias edgar.igles...@xilinx.com
Signed-off-by: Edgar E. Iglesias edgar.igles...@xilinx.com
---
cputlb.c| 2 +-
exec.c | 5 ++---
include/exec/exec-all.h | 2 +-
include/exec/softmmu_template.h | 5 +++--
4 files changed, 7 insertions(+), 7 deletions(-)
diff --git a/cputlb.c b/cputlb.c
index b533f3f..71cf64f 100644
--- a/cputlb.c
+++ b/cputlb.c
@@ -313,7 +313,7 @@ tb_page_addr_t get_page_addr_code(CPUArchState *env1,
target_ulong addr)
cpu_ldub_code(env1, addr);
}
pd = env1-iotlb[mmu_idx][page_index] ~TARGET_PAGE_MASK;
-mr = iotlb_to_region(pd);
+mr = iotlb_to_region(address_space_memory, pd);
if (memory_region_is_unassigned(mr)) {
CPUState *cpu = ENV_GET_CPU(env1);
CPUClass *cc = CPU_GET_CLASS(cpu);
diff --git a/exec.c b/exec.c
index fc8ae65..8a90867 100644
--- a/exec.c
+++ b/exec.c
@@ -1731,10 +1731,9 @@ static uint16_t dummy_section(PhysPageMap *map,
MemoryRegion *mr)
return phys_section_add(map, section);
}
-MemoryRegion *iotlb_to_region(hwaddr index)
+MemoryRegion *iotlb_to_region(AddressSpace *as, hwaddr index)
{
-return address_space_memory.dispatch-map.sections[
- index ~TARGET_PAGE_MASK].mr;
+return as-dispatch-map.sections[index ~TARGET_PAGE_MASK].mr;
}
static void io_mem_init(void)
diff --git a/include/exec/exec-all.h b/include/exec/exec-all.h
index 9d6c36b..75fd1da 100644
--- a/include/exec/exec-all.h
+++ b/include/exec/exec-all.h
@@ -325,7 +325,7 @@ extern uintptr_t tci_tb_ptr;
void phys_mem_set_alloc(void *(*alloc)(size_t));
-struct MemoryRegion *iotlb_to_region(hwaddr index);
+struct MemoryRegion *iotlb_to_region(AddressSpace *as, hwaddr index);
bool io_mem_read(struct MemoryRegion *mr, hwaddr addr,
uint64_t *pvalue, unsigned size);
bool io_mem_write(struct MemoryRegion *mr, hwaddr addr,
diff --git a/include/exec/softmmu_template.h b/include/exec/softmmu_template.h
index c6a5440..69d856a 100644
--- a/include/exec/softmmu_template.h
+++ b/include/exec/softmmu_template.h
@@ -22,6 +22,7 @@
* License along with this library; if not, see http://www.gnu.org/licenses/.
*/
#include qemu/timer.h
+#include exec/address-spaces.h
#include exec/memory.h
#define DATA_SIZE (1 SHIFT)
@@ -118,7 +119,7 @@ static inline DATA_TYPE glue(io_read, SUFFIX)(CPUArchState
*env,
uintptr_t retaddr)
{
uint64_t val;
-MemoryRegion *mr = iotlb_to_region(physaddr);
+MemoryRegion *mr = iotlb_to_region(address_space_memory, physaddr);
physaddr = (physaddr TARGET_PAGE_MASK) + addr;
env-mem_io_pc = retaddr;
@@ -324,7 +325,7 @@ static inline void glue(io_write, SUFFIX)(CPUArchState *env,
target_ulong addr,
uintptr_t retaddr)
{
-MemoryRegion *mr = iotlb_to_region(physaddr);
+MemoryRegion *mr = iotlb_to_region(address_space_memory, physaddr);
physaddr = (physaddr TARGET_PAGE_MASK) + addr;
if (mr != io_mem_rom mr != io_mem_notdirty !can_do_io(env)) {
--
1.8.1.2
[Qemu-devel] [PATCH v4 04/22] exec: Make memory_region_section_get_iotlb use section AS
From: Edgar E. Iglesias edgar.igles...@xilinx.com
Signed-off-by: Edgar E. Iglesias edgar.igles...@xilinx.com
---
exec.c | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
diff --git a/exec.c b/exec.c
index 3814a7e..d8cfe52 100644
--- a/exec.c
+++ b/exec.c
@@ -781,7 +781,7 @@ hwaddr memory_region_section_get_iotlb(CPUArchState *env,
iotlb |= PHYS_SECTION_ROM;
}
} else {
-iotlb = section - address_space_memory.dispatch-map.sections;
+iotlb = section - section-address_space-dispatch-map.sections;
iotlb += xlat;
}
--
1.8.1.2
Re: [Qemu-devel] [PATCH 20/42] input: mouse: add graphic_rotate support
Il 03/02/2014 10:42, Gerd Hoffmann ha scritto:
Hi,
+switch (graphic_rotate) {
+case 90:
+if (evt-abs-axis == INPUT_AXIS_X) {
+evt-abs-axis = INPUT_AXIS_Y;
+}
+if (evt-abs-axis == INPUT_AXIS_Y) {
Need else here, same for case 270.
Why?
Because otherwise both axes become X. When the first if triggers the
axis changes to Y, and then it becomes X again. :)
+evt-abs-axis = INPUT_AXIS_X;
+evt-abs-axis = INPUT_EVENT_ABS_SIZE - 1 - evt-abs-axis;
-value here, not -axis.
Oops, indeed.
It looks like doing it right for relative is easy:
But what is the point when this isn't used anyway?
In what sense it is not used? Because you do not have e.g. a USB
controller in the only machines that support graphic_rotate?
Paolo
if (graphic_rotate == 0) {
return;
}
if (move-axis != INPUT_AXIS_X move-axis != INPUT_AXIS_Y) {
return;
}
if ((graphic_rotate = 180 move-axis == INPUT_AXIS_X) ||
(graphic_rotate = 180 move-axis == INPUT_AXIS_Y)) {
if (kind == INPUT_EVENT_KIND_ABS) {
move-value = INPUT_EVENT_ABS_SIZE - 1 - move-value;
} else {
move-value = -move-value;
}
}
if (graphic_rotate == 90 || graphic_rotate == 270) {
move-axis ^= INPUT_AXIS_X ^ INPUT_AXIS_Y;
}
Saves a few lines but it's a bit harder to figure what is going on ...
cheers,
Gerd
[Qemu-devel] [PATCH v4 00/22] Steps towards per CPU address-spaces
From: Edgar E. Iglesias edgar.igles...@xilinx.com Hi, I'm looking at modeling systems where multiple CPUs co-exist with different views of their attached buses/devs. With this series I'm trying to take some steps towards having an address-space per CPU. It's not complete but good enough for making it possible to model (to some extent) CPU local memories for MicroBlaze systems in emulation mode (TCG). I'm updating the petalogix-ml605 here and will follow-up later with the petalogix-s3adsp. The per-cpu address space is added into the CPUState. I tried to measure performance diff with having it in the CPUState-env. For normal and even for IO heavy workloads on linux kernels, the diff is not measurable. I also tested with a tight guest loop that continuously does I/O accesses and there I can see a 2.5% drop in perf. I dont think the runtime type check involved when casting from env to CS will be much of a problem. I've reordered the series and moved the AS props to the end, hoping we can get through the bulk of the series with less controversy and get it commited soon. I've kept the interface with properties to set AddressSpace pointers which I think is the more flexible approach but we can explore other ideas if there are. There is lots of future work needed, for example to transform more of the cpu_* bus accessing functions. To add more usage of AddressSpace properties to pass on address spaces to DMA models. Qtest mechanisms to target specific address spaces, etc... Cheers, Edgar v3 - v4: Rebase, Use error_abort. v2 - v3: Move CPU address-space prop into CPUState level. v1 - v2: Add braces in cpu_memory_rw_debug. Avoid mixing var/code declarations in tcg_commit. Move per-cpu address space into CPUState. Reorder patch series to add the AS properties last. Edgar E. Iglesias (22): exec: Make tb_invalidate_phys_addr input an AS exec: Make iotlb_to_region input an AS exec: Always initialize MemorySection address spaces exec: Make memory_region_section_get_iotlb use section AS memory: Add MemoryListener to typedefs.h cpu: Add per-cpu address space exec: On AS changes, only flush affected CPU TLBs exec: Make ldl_*_phys input an AddressSpace exec: Make ldq/ldub_*_phys input an AddressSpace exec: Make lduw_*_phys input an AddressSpace exec: Make stq_*_phys input an AddressSpace exec: Make stl_*_phys input an AddressSpace exec: Make stl_phys_notdirty input an AddressSpace exec: Make stw_*_phys input an AddressSpace exec: Make stb_phys input an AddressSpace exec: Make cpu_physical_memory_write_rom input an AS exec: Make cpu_memory_rw_debug use the CPUs AS memory: Add address_space_find_by_name() qdev: Add qdev property type for AddressSpaces cpu: Add address-space property petalogix-ml605: Create the CPU with object_new() petalogix-ml605: Make the LMB visible only to the CPU cpu-exec.c | 5 +- cpus.c | 2 + cputlb.c| 7 +- exec.c | 183 -- hw/alpha/dp264.c| 5 +- hw/alpha/typhoon.c | 2 +- hw/arm/boot.c | 5 +- hw/arm/highbank.c | 6 +- hw/core/loader.c| 3 +- hw/core/qdev-properties-system.c| 8 + hw/core/qdev-properties.c | 54 +++ hw/display/sm501.c | 1 + hw/display/sm501_template.h | 2 +- hw/dma/pl080.c | 9 +- hw/dma/sun4m_iommu.c| 3 +- hw/intc/apic.c | 3 +- hw/microblaze/petalogix_ml605_mmu.c | 24 ++- hw/net/vmware_utils.h | 16 +- hw/pci/msi.c| 2 +- hw/pci/msix.c | 2 +- hw/ppc/ppc405_uc.c | 45 +++--- hw/ppc/spapr_hcall.c| 50 +++--- hw/s390x/css.c | 11 +- hw/s390x/s390-virtio-bus.c | 36 +++-- hw/s390x/s390-virtio.c | 2 +- hw/s390x/virtio-ccw.c | 40 +++-- hw/scsi/megasas.c | 22 ++- hw/scsi/vmw_pvscsi.c| 6 +- hw/sh4/r2d.c| 4 +- hw/sparc/sun4m.c| 3 +- hw/timer/hpet.c | 3 +- hw/virtio/virtio.c | 31 ++-- include/exec/cpu-common.h | 44 +++--- include/exec/exec-all.h | 5 +- include/exec/memory.h | 11 +- include/exec/softmmu_template.h | 7 +- include/hw/ppc/spapr.h | 4 +- include/hw/qdev-properties.h| 5 + include/qemu/typedefs.h | 1 + include/qom/cpu.h | 3 + memory.c| 12 ++ monitor.c | 2 +- qom/cpu.c | 7 + stubs/Makefile.objs | 1 + stubs/memory.c | 6 +
[Qemu-devel] [PATCH v4 05/22] memory: Add MemoryListener to typedefs.h
From: Edgar E. Iglesias edgar.igles...@xilinx.com
Signed-off-by: Edgar E. Iglesias edgar.igles...@xilinx.com
---
include/exec/memory.h | 2 --
include/qemu/typedefs.h | 1 +
2 files changed, 1 insertion(+), 2 deletions(-)
diff --git a/include/exec/memory.h b/include/exec/memory.h
index 296d6ab..9101fc3 100644
--- a/include/exec/memory.h
+++ b/include/exec/memory.h
@@ -163,8 +163,6 @@ struct MemoryRegion {
NotifierList iommu_notify;
};
-typedef struct MemoryListener MemoryListener;
-
/**
* MemoryListener: callbacks structure for updates to the physical memory map
*
diff --git a/include/qemu/typedefs.h b/include/qemu/typedefs.h
index 4524496..5b4e333 100644
--- a/include/qemu/typedefs.h
+++ b/include/qemu/typedefs.h
@@ -26,6 +26,7 @@ typedef struct BusClass BusClass;
typedef struct AddressSpace AddressSpace;
typedef struct MemoryRegion MemoryRegion;
typedef struct MemoryRegionSection MemoryRegionSection;
+typedef struct MemoryListener MemoryListener;
typedef struct MemoryMappingList MemoryMappingList;
--
1.8.1.2
[Qemu-devel] [PATCH v4 07/22] exec: On AS changes, only flush affected CPU TLBs
From: Edgar E. Iglesias edgar.igles...@xilinx.com
Signed-off-by: Edgar E. Iglesias edgar.igles...@xilinx.com
---
exec.c | 5 +
1 file changed, 5 insertions(+)
diff --git a/exec.c b/exec.c
index f299d06..a1d8bbe 100644
--- a/exec.c
+++ b/exec.c
@@ -1813,6 +1813,11 @@ static void tcg_commit(MemoryListener *listener)
CPU_FOREACH(cpu) {
CPUArchState *env = cpu-env_ptr;
+/* FIXME: Disentangle the cpu.h circular files deps so we can
+ directly get the right CPU from listener. */
+if (cpu-tcg_as_listener != listener) {
+continue;
+}
tlb_flush(env, 1);
}
}
--
1.8.1.2
[Qemu-devel] [PATCH v4 10/22] exec: Make lduw_*_phys input an AddressSpace
From: Edgar E. Iglesias edgar.igles...@xilinx.com
Signed-off-by: Edgar E. Iglesias edgar.igles...@xilinx.com
---
exec.c | 18 +-
hw/net/vmware_utils.h | 2 +-
hw/ppc/spapr_hcall.c | 6 +++---
hw/s390x/css.c | 3 ++-
hw/s390x/virtio-ccw.c | 8 +---
hw/virtio/virtio.c | 16
include/exec/cpu-common.h | 6 +++---
target-i386/smm_helper.c | 12 ++--
target-i386/svm_helper.c | 15 ---
target-sparc/ldst_helper.c | 6 +++---
10 files changed, 48 insertions(+), 44 deletions(-)
diff --git a/exec.c b/exec.c
index 3bc4497..11abe89 100644
--- a/exec.c
+++ b/exec.c
@@ -1598,7 +1598,7 @@ static uint64_t watch_mem_read(void *opaque, hwaddr addr,
check_watchpoint(addr ~TARGET_PAGE_MASK, ~(size - 1), BP_MEM_READ);
switch (size) {
case 1: return ldub_phys(address_space_memory, addr);
-case 2: return lduw_phys(addr);
+case 2: return lduw_phys(address_space_memory, addr);
case 4: return ldl_phys(address_space_memory, addr);
default: abort();
}
@@ -2470,7 +2470,7 @@ uint32_t ldub_phys(AddressSpace *as, hwaddr addr)
}
/* warning: addr must be aligned */
-static inline uint32_t lduw_phys_internal(hwaddr addr,
+static inline uint32_t lduw_phys_internal(AddressSpace *as, hwaddr addr,
enum device_endian endian)
{
uint8_t *ptr;
@@ -2479,7 +2479,7 @@ static inline uint32_t lduw_phys_internal(hwaddr addr,
hwaddr l = 2;
hwaddr addr1;
-mr = address_space_translate(address_space_memory, addr, addr1, l,
+mr = address_space_translate(as, addr, addr1, l,
false);
if (l 2 || !memory_access_is_direct(mr, false)) {
/* I/O case */
@@ -2513,19 +2513,19 @@ static inline uint32_t lduw_phys_internal(hwaddr addr,
return val;
}
-uint32_t lduw_phys(hwaddr addr)
+uint32_t lduw_phys(AddressSpace *as, hwaddr addr)
{
-return lduw_phys_internal(addr, DEVICE_NATIVE_ENDIAN);
+return lduw_phys_internal(as, addr, DEVICE_NATIVE_ENDIAN);
}
-uint32_t lduw_le_phys(hwaddr addr)
+uint32_t lduw_le_phys(AddressSpace *as, hwaddr addr)
{
-return lduw_phys_internal(addr, DEVICE_LITTLE_ENDIAN);
+return lduw_phys_internal(as, addr, DEVICE_LITTLE_ENDIAN);
}
-uint32_t lduw_be_phys(hwaddr addr)
+uint32_t lduw_be_phys(AddressSpace *as, hwaddr addr)
{
-return lduw_phys_internal(addr, DEVICE_BIG_ENDIAN);
+return lduw_phys_internal(as, addr, DEVICE_BIG_ENDIAN);
}
/* warning: addr must be aligned. The ram page is not masked as dirty
diff --git a/hw/net/vmware_utils.h b/hw/net/vmware_utils.h
index eb98d2a..7877cb6 100644
--- a/hw/net/vmware_utils.h
+++ b/hw/net/vmware_utils.h
@@ -80,7 +80,7 @@ vmw_shmem_st8(hwaddr addr, uint8_t value)
static inline uint32_t
vmw_shmem_ld16(hwaddr addr)
{
-uint16_t res = lduw_le_phys(addr);
+uint16_t res = lduw_le_phys(address_space_memory, addr);
VMW_SHPRN(SHMEM load16: % PRIx64 (value 0x%X), addr, res);
return res;
}
diff --git a/hw/ppc/spapr_hcall.c b/hw/ppc/spapr_hcall.c
index 6e61a91..a2893ad 100644
--- a/hw/ppc/spapr_hcall.c
+++ b/hw/ppc/spapr_hcall.c
@@ -355,7 +355,7 @@ static target_ulong register_vpa(CPUPPCState *env,
target_ulong vpa)
}
/* FIXME: bounds check the address */
-size = lduw_be_phys(vpa + 0x4);
+size = lduw_be_phys(cs-as, vpa + 0x4);
if (size VPA_MIN_SIZE) {
return H_PARAMETER;
@@ -544,7 +544,7 @@ static target_ulong h_logical_load(PowerPCCPU *cpu,
sPAPREnvironment *spapr,
args[0] = ldub_phys(cs-as, addr);
return H_SUCCESS;
case 2:
-args[0] = lduw_phys(addr);
+args[0] = lduw_phys(cs-as, addr);
return H_SUCCESS;
case 4:
args[0] = ldl_phys(cs-as, addr);
@@ -614,7 +614,7 @@ static target_ulong h_logical_memop(PowerPCCPU *cpu,
sPAPREnvironment *spapr,
tmp = ldub_phys(cs-as, src);
break;
case 1:
-tmp = lduw_phys(src);
+tmp = lduw_phys(cs-as, src);
break;
case 2:
tmp = ldl_phys(cs-as, src);
diff --git a/hw/s390x/css.c b/hw/s390x/css.c
index 41b1903..d42d7ec 100644
--- a/hw/s390x/css.c
+++ b/hw/s390x/css.c
@@ -677,7 +677,8 @@ static void css_update_chnmon(SubchDev *sch)
uint16_t count;
offset = sch-curr_status.pmcw.mbi 5;
-count = lduw_phys(channel_subsys-chnmon_area + offset);
+count = lduw_phys(address_space_memory,
+ channel_subsys-chnmon_area + offset);
count++;
stw_phys(channel_subsys-chnmon_area + offset, count);
}
diff --git a/hw/s390x/virtio-ccw.c b/hw/s390x/virtio-ccw.c
index d301f00..50d852a 100644
--- a/hw/s390x/virtio-ccw.c
+++ b/hw/s390x/virtio-ccw.c
@@ -265,9 +265,11 @@ static int virtio_ccw_cb(SubchDev *sch, CCW1 ccw)
info.queue = ldq_phys(address_space_memory, ccw.cda);
[Qemu-devel] [PATCH v4 06/22] cpu: Add per-cpu address space
From: Edgar E. Iglesias edgar.igles...@xilinx.com
Signed-off-by: Edgar E. Iglesias edgar.igles...@xilinx.com
---
cpus.c | 2 ++
cputlb.c| 7 ---
exec.c | 27 +++
include/exec/exec-all.h | 1 +
include/exec/softmmu_template.h | 6 --
include/qom/cpu.h | 3 +++
6 files changed, 33 insertions(+), 13 deletions(-)
diff --git a/cpus.c b/cpus.c
index ca4c59f..945d85b 100644
--- a/cpus.c
+++ b/cpus.c
@@ -1119,6 +1119,8 @@ void resume_all_vcpus(void)
static void qemu_tcg_init_vcpu(CPUState *cpu)
{
+tcg_cpu_address_space_init(cpu, cpu-as);
+
/* share a single thread for all cpus with TCG */
if (!tcg_cpu_thread) {
cpu-thread = g_malloc0(sizeof(QemuThread));
diff --git a/cputlb.c b/cputlb.c
index 71cf64f..0fbaa39 100644
--- a/cputlb.c
+++ b/cputlb.c
@@ -232,6 +232,7 @@ void tlb_set_page(CPUArchState *env, target_ulong vaddr,
uintptr_t addend;
CPUTLBEntry *te;
hwaddr iotlb, xlat, sz;
+CPUState *cpu = ENV_GET_CPU(env);
assert(size = TARGET_PAGE_SIZE);
if (size != TARGET_PAGE_SIZE) {
@@ -239,7 +240,7 @@ void tlb_set_page(CPUArchState *env, target_ulong vaddr,
}
sz = size;
-section = address_space_translate_for_iotlb(address_space_memory, paddr,
+section = address_space_translate_for_iotlb(cpu-as, paddr,
xlat, sz);
assert(sz = TARGET_PAGE_SIZE);
@@ -305,6 +306,7 @@ tb_page_addr_t get_page_addr_code(CPUArchState *env1,
target_ulong addr)
int mmu_idx, page_index, pd;
void *p;
MemoryRegion *mr;
+CPUState *cpu = ENV_GET_CPU(env1);
page_index = (addr TARGET_PAGE_BITS) (CPU_TLB_SIZE - 1);
mmu_idx = cpu_mmu_index(env1);
@@ -313,9 +315,8 @@ tb_page_addr_t get_page_addr_code(CPUArchState *env1,
target_ulong addr)
cpu_ldub_code(env1, addr);
}
pd = env1-iotlb[mmu_idx][page_index] ~TARGET_PAGE_MASK;
-mr = iotlb_to_region(address_space_memory, pd);
+mr = iotlb_to_region(cpu-as, pd);
if (memory_region_is_unassigned(mr)) {
-CPUState *cpu = ENV_GET_CPU(env1);
CPUClass *cc = CPU_GET_CLASS(cpu);
if (cc-do_unassigned_access) {
diff --git a/exec.c b/exec.c
index d8cfe52..f299d06 100644
--- a/exec.c
+++ b/exec.c
@@ -138,6 +138,7 @@ typedef struct subpage_t {
static void io_mem_init(void);
static void memory_map_init(void);
+static void tcg_commit(MemoryListener *listener);
static MemoryRegion io_mem_watch;
#endif
@@ -438,6 +439,22 @@ CPUState *qemu_get_cpu(int index)
return NULL;
}
+#if !defined(CONFIG_USER_ONLY)
+void tcg_cpu_address_space_init(CPUState *cpu, AddressSpace *as)
+{
+/* We only support one address space per cpu at the moment. */
+assert(cpu-as == as);
+
+if (cpu-tcg_as_listener) {
+memory_listener_unregister(cpu-tcg_as_listener);
+} else {
+cpu-tcg_as_listener = g_new0(MemoryListener, 1);
+}
+cpu-tcg_as_listener-commit = tcg_commit;
+memory_listener_register(cpu-tcg_as_listener, as);
+}
+#endif
+
void cpu_exec_init(CPUArchState *env)
{
CPUState *cpu = ENV_GET_CPU(env);
@@ -457,6 +474,7 @@ void cpu_exec_init(CPUArchState *env)
QTAILQ_INIT(env-breakpoints);
QTAILQ_INIT(env-watchpoints);
#ifndef CONFIG_USER_ONLY
+cpu-as = address_space_memory;
cpu-thread_id = qemu_get_thread_id();
#endif
QTAILQ_INSERT_TAIL(cpus, cpu, node);
@@ -488,7 +506,7 @@ static void breakpoint_invalidate(CPUState *cpu,
target_ulong pc)
{
hwaddr phys = cpu_get_phys_page_debug(cpu, pc);
if (phys != -1) {
-tb_invalidate_phys_addr(address_space_memory,
+tb_invalidate_phys_addr(cpu-as,
phys | (pc ~TARGET_PAGE_MASK));
}
}
@@ -1815,10 +1833,6 @@ static MemoryListener core_memory_listener = {
.priority = 1,
};
-static MemoryListener tcg_memory_listener = {
-.commit = tcg_commit,
-};
-
void address_space_init_dispatch(AddressSpace *as)
{
as-dispatch = NULL;
@@ -1854,9 +1868,6 @@ static void memory_map_init(void)
address_space_init(address_space_io, system_io, I/O);
memory_listener_register(core_memory_listener, address_space_memory);
-if (tcg_enabled()) {
-memory_listener_register(tcg_memory_listener, address_space_memory);
-}
}
MemoryRegion *get_system_memory(void)
diff --git a/include/exec/exec-all.h b/include/exec/exec-all.h
index 75fd1da..a387922 100644
--- a/include/exec/exec-all.h
+++ b/include/exec/exec-all.h
@@ -96,6 +96,7 @@ void tb_invalidate_phys_page_range(tb_page_addr_t start,
tb_page_addr_t end,
void tb_invalidate_phys_range(tb_page_addr_t start, tb_page_addr_t end,
int is_cpu_write_access);
#if !defined(CONFIG_USER_ONLY)
+void tcg_cpu_address_space_init(CPUState *cpu, AddressSpace *as);
/* cputlb.c */
void tlb_flush_page(CPUArchState *env,
[Qemu-devel] [PATCH v4 08/22] exec: Make ldl_*_phys input an AddressSpace
From: Edgar E. Iglesias edgar.igles...@xilinx.com
Signed-off-by: Edgar E. Iglesias edgar.igles...@xilinx.com
---
cpu-exec.c| 5 +-
exec.c| 19 ---
hw/dma/pl080.c| 9 ++--
hw/dma/sun4m_iommu.c | 3 +-
hw/net/vmware_utils.h | 2 +-
hw/ppc/spapr_hcall.c | 13 +++--
hw/s390x/css.c| 3 +-
hw/s390x/s390-virtio-bus.c| 2 +-
hw/s390x/virtio-ccw.c | 5 +-
hw/scsi/megasas.c | 4 +-
hw/scsi/vmw_pvscsi.c | 3 +-
hw/virtio/virtio.c| 3 +-
include/exec/cpu-common.h | 6 +--
include/hw/ppc/spapr.h| 2 +-
target-alpha/helper.h | 2 +-
target-alpha/mem_helper.c | 11 ++--
target-alpha/translate.c | 2 +-
target-arm/helper.c | 15 +++---
target-i386/arch_memory_mapping.c | 36 ++---
target-i386/helper.c | 9 ++--
target-i386/seg_helper.c | 6 ++-
target-i386/smm_helper.c | 103 --
target-i386/svm_helper.c | 28 ++-
target-ppc/excp_helper.c | 4 +-
target-ppc/mmu-hash32.h | 6 ++-
target-s390x/cpu.c| 2 +-
target-sparc/ldst_helper.c| 8 +--
target-sparc/mmu_helper.c | 20
target-unicore32/softmmu.c| 5 +-
target-xtensa/helper.c| 3 +-
30 files changed, 190 insertions(+), 149 deletions(-)
diff --git a/cpu-exec.c b/cpu-exec.c
index a6c01f4..8943493 100644
--- a/cpu-exec.c
+++ b/cpu-exec.c
@@ -395,7 +395,10 @@ int cpu_exec(CPUArchState *env)
/* FIXME: this should respect TPR */
cpu_svm_check_intercept_param(env, SVM_EXIT_VINTR,
0);
-intno = ldl_phys(env-vm_vmcb + offsetof(struct
vmcb, control.int_vector));
+intno = ldl_phys(cpu-as,
+ env-vm_vmcb
+ + offsetof(struct vmcb,
+control.int_vector));
qemu_log_mask(CPU_LOG_TB_IN_ASM, Servicing
virtual hardware INT=0x%02x\n, intno);
do_interrupt_x86_hardirq(env, intno, 1);
cpu-interrupt_request = ~CPU_INTERRUPT_VIRQ;
diff --git a/exec.c b/exec.c
index a1d8bbe..aa208be 100644
--- a/exec.c
+++ b/exec.c
@@ -1599,7 +1599,7 @@ static uint64_t watch_mem_read(void *opaque, hwaddr addr,
switch (size) {
case 1: return ldub_phys(addr);
case 2: return lduw_phys(addr);
-case 4: return ldl_phys(addr);
+case 4: return ldl_phys(address_space_memory, addr);
default: abort();
}
}
@@ -2345,7 +2345,7 @@ void cpu_physical_memory_unmap(void *buffer, hwaddr len,
}
/* warning: addr must be aligned */
-static inline uint32_t ldl_phys_internal(hwaddr addr,
+static inline uint32_t ldl_phys_internal(AddressSpace *as, hwaddr addr,
enum device_endian endian)
{
uint8_t *ptr;
@@ -2354,8 +2354,7 @@ static inline uint32_t ldl_phys_internal(hwaddr addr,
hwaddr l = 4;
hwaddr addr1;
-mr = address_space_translate(address_space_memory, addr, addr1, l,
- false);
+mr = address_space_translate(as, addr, addr1, l, false);
if (l 4 || !memory_access_is_direct(mr, false)) {
/* I/O case */
io_mem_read(mr, addr1, val, 4);
@@ -2388,19 +2387,19 @@ static inline uint32_t ldl_phys_internal(hwaddr addr,
return val;
}
-uint32_t ldl_phys(hwaddr addr)
+uint32_t ldl_phys(AddressSpace *as, hwaddr addr)
{
-return ldl_phys_internal(addr, DEVICE_NATIVE_ENDIAN);
+return ldl_phys_internal(as, addr, DEVICE_NATIVE_ENDIAN);
}
-uint32_t ldl_le_phys(hwaddr addr)
+uint32_t ldl_le_phys(AddressSpace *as, hwaddr addr)
{
-return ldl_phys_internal(addr, DEVICE_LITTLE_ENDIAN);
+return ldl_phys_internal(as, addr, DEVICE_LITTLE_ENDIAN);
}
-uint32_t ldl_be_phys(hwaddr addr)
+uint32_t ldl_be_phys(AddressSpace *as, hwaddr addr)
{
-return ldl_phys_internal(addr, DEVICE_BIG_ENDIAN);
+return ldl_phys_internal(as, addr, DEVICE_BIG_ENDIAN);
}
/* warning: addr must be aligned */
diff --git a/hw/dma/pl080.c b/hw/dma/pl080.c
index cb7bda9..741dd20 100644
--- a/hw/dma/pl080.c
+++ b/hw/dma/pl080.c
@@ -8,6 +8,7 @@
*/
#include hw/sysbus.h
+#include exec/address-spaces.h
#define PL080_MAX_CHANNELS 8
#define PL080_CONF_E0x1
@@ -204,10 +205,10 @@ again:
if (size == 0) {
/* Transfer complete. */
if (ch-lli) {
-ch-src = ldl_le_phys(ch-lli);
-ch-dest = ldl_le_phys(ch-lli + 4);
-
[Qemu-devel] [PATCH v4 12/22] exec: Make stl_*_phys input an AddressSpace
From: Edgar E. Iglesias edgar.igles...@xilinx.com
Signed-off-by: Edgar E. Iglesias edgar.igles...@xilinx.com
---
exec.c | 19 +
hw/net/vmware_utils.h | 2 +-
hw/pci/msi.c | 2 +-
hw/pci/msix.c | 2 +-
hw/ppc/ppc405_uc.c | 33 +++
hw/ppc/spapr_hcall.c | 4 +-
hw/s390x/css.c | 2 +-
hw/s390x/s390-virtio-bus.c | 2 +-
hw/s390x/virtio-ccw.c | 2 +-
hw/scsi/megasas.c | 6 ++-
hw/scsi/vmw_pvscsi.c | 3 +-
hw/sh4/r2d.c | 2 +-
hw/timer/hpet.c| 3 +-
hw/virtio/virtio.c | 4 +-
include/exec/cpu-common.h | 6 +--
include/hw/ppc/spapr.h | 2 +-
target-alpha/helper.h | 2 +-
target-alpha/mem_helper.c | 7 ++--
target-alpha/translate.c | 2 +-
target-arm/helper.c| 3 +-
target-i386/seg_helper.c | 8 ++--
target-i386/smm_helper.c | 100 ++---
target-i386/svm_helper.c | 28 -
target-ppc/mmu-hash32.h| 6 ++-
target-sparc/ldst_helper.c | 6 +--
25 files changed, 137 insertions(+), 119 deletions(-)
diff --git a/exec.c b/exec.c
index f8be6da..a1f720c 100644
--- a/exec.c
+++ b/exec.c
@@ -1616,7 +1616,7 @@ static void watch_mem_write(void *opaque, hwaddr addr,
stw_phys(addr, val);
break;
case 4:
-stl_phys(addr, val);
+stl_phys(address_space_memory, addr, val);
break;
default: abort();
}
@@ -2561,7 +2561,8 @@ void stl_phys_notdirty(hwaddr addr, uint32_t val)
}
/* warning: addr must be aligned */
-static inline void stl_phys_internal(hwaddr addr, uint32_t val,
+static inline void stl_phys_internal(AddressSpace *as,
+ hwaddr addr, uint32_t val,
enum device_endian endian)
{
uint8_t *ptr;
@@ -2569,7 +2570,7 @@ static inline void stl_phys_internal(hwaddr addr,
uint32_t val,
hwaddr l = 4;
hwaddr addr1;
-mr = address_space_translate(address_space_memory, addr, addr1, l,
+mr = address_space_translate(as, addr, addr1, l,
true);
if (l 4 || !memory_access_is_direct(mr, true)) {
#if defined(TARGET_WORDS_BIGENDIAN)
@@ -2601,19 +2602,19 @@ static inline void stl_phys_internal(hwaddr addr,
uint32_t val,
}
}
-void stl_phys(hwaddr addr, uint32_t val)
+void stl_phys(AddressSpace *as, hwaddr addr, uint32_t val)
{
-stl_phys_internal(addr, val, DEVICE_NATIVE_ENDIAN);
+stl_phys_internal(as, addr, val, DEVICE_NATIVE_ENDIAN);
}
-void stl_le_phys(hwaddr addr, uint32_t val)
+void stl_le_phys(AddressSpace *as, hwaddr addr, uint32_t val)
{
-stl_phys_internal(addr, val, DEVICE_LITTLE_ENDIAN);
+stl_phys_internal(as, addr, val, DEVICE_LITTLE_ENDIAN);
}
-void stl_be_phys(hwaddr addr, uint32_t val)
+void stl_be_phys(AddressSpace *as, hwaddr addr, uint32_t val)
{
-stl_phys_internal(addr, val, DEVICE_BIG_ENDIAN);
+stl_phys_internal(as, addr, val, DEVICE_BIG_ENDIAN);
}
/* XXX: optimize */
diff --git a/hw/net/vmware_utils.h b/hw/net/vmware_utils.h
index 4cf0e79..2ed73af 100644
--- a/hw/net/vmware_utils.h
+++ b/hw/net/vmware_utils.h
@@ -104,7 +104,7 @@ static inline void
vmw_shmem_st32(hwaddr addr, uint32_t value)
{
VMW_SHPRN(SHMEM store32: % PRIx64 (value 0x%X), addr, value);
-stl_le_phys(addr, value);
+stl_le_phys(address_space_memory, addr, value);
}
static inline uint64_t
diff --git a/hw/pci/msi.c b/hw/pci/msi.c
index 2a04d18..a4a3040 100644
--- a/hw/pci/msi.c
+++ b/hw/pci/msi.c
@@ -291,7 +291,7 @@ void msi_notify(PCIDevice *dev, unsigned int vector)
notify vector 0x%x
address: 0x%PRIx64 data: 0x%PRIx32\n,
vector, msg.address, msg.data);
-stl_le_phys(msg.address, msg.data);
+stl_le_phys(address_space_memory, msg.address, msg.data);
}
/* Normally called by pci_default_write_config(). */
diff --git a/hw/pci/msix.c b/hw/pci/msix.c
index 3430770..5c49bfc 100644
--- a/hw/pci/msix.c
+++ b/hw/pci/msix.c
@@ -439,7 +439,7 @@ void msix_notify(PCIDevice *dev, unsigned vector)
msg = msix_get_message(dev, vector);
-stl_le_phys(msg.address, msg.data);
+stl_le_phys(address_space_memory, msg.address, msg.data);
}
void msix_reset(PCIDevice *dev)
diff --git a/hw/ppc/ppc405_uc.c b/hw/ppc/ppc405_uc.c
index 8109f92..47a4242 100644
--- a/hw/ppc/ppc405_uc.c
+++ b/hw/ppc/ppc405_uc.c
@@ -44,6 +44,7 @@
ram_addr_t ppc405_set_bootinfo (CPUPPCState *env, ppc4xx_bd_info_t *bd,
uint32_t flags)
{
+CPUState *cs = ENV_GET_CPU(env);
ram_addr_t bdloc;
int i, n;
@@ -52,30 +53,30 @@ ram_addr_t ppc405_set_bootinfo (CPUPPCState *env,
ppc4xx_bd_info_t *bd,
bdloc = 0x0100UL - sizeof(struct ppc4xx_bd_info_t);
else
bdloc = bd-bi_memsize - sizeof(struct ppc4xx_bd_info_t);
-
Re: [Qemu-devel] [PATCH 01/10] block: Change BDS parameter of bdrv_open() to **
Am 31.01.2014 um 21:07 hat Max Reitz geschrieben: So in summary, it looks technically correct, but I'm not sure about maintainability. If the problems are only for this intermediate state, I'm okay with leaving them as they are, but otherwise we should try and figure out something. You're right in that bdrv_open should always create a new BDS and new BDS should always be created through bdrv_open; but right now, there are some places in the code (such as blockdev_init()) which I was just afraid of touching as they seemed to heavily rely on being able to separate those two steps. Sure, there's a good reason why we work in incremental steps. This is a huge conversion of infrastructure and trying to change it all at once never worked out. We shouldn't worry too much about ugly intermediate states as long as we're sure that they are really just intermediate. Kevin
[Qemu-devel] [PATCH v4 11/22] exec: Make stq_*_phys input an AddressSpace
From: Edgar E. Iglesias edgar.igles...@xilinx.com
Signed-off-by: Edgar E. Iglesias edgar.igles...@xilinx.com
---
exec.c | 12 +++---
hw/alpha/dp264.c | 5 ++-
hw/net/vmware_utils.h | 2 +-
hw/ppc/spapr_hcall.c | 6 ++-
hw/s390x/s390-virtio-bus.c | 3 +-
hw/s390x/virtio-ccw.c | 4 +-
hw/scsi/megasas.c | 3 +-
include/exec/cpu-common.h | 6 +--
target-alpha/helper.h | 2 +-
target-alpha/mem_helper.c | 7 +--
target-alpha/translate.c | 2 +-
target-i386/helper.c | 3 +-
target-i386/smm_helper.c | 33 ---
target-i386/svm_helper.c | 103 +++--
target-ppc/mmu-hash64.h| 7 ++-
target-s390x/helper.c | 8 +++-
target-s390x/mem_helper.c | 3 +-
target-sparc/ldst_helper.c | 15 ---
18 files changed, 132 insertions(+), 92 deletions(-)
diff --git a/exec.c b/exec.c
index 11abe89..f8be6da 100644
--- a/exec.c
+++ b/exec.c
@@ -2680,22 +2680,22 @@ void stw_be_phys(hwaddr addr, uint32_t val)
}
/* XXX: optimize */
-void stq_phys(hwaddr addr, uint64_t val)
+void stq_phys(AddressSpace *as, hwaddr addr, uint64_t val)
{
val = tswap64(val);
-cpu_physical_memory_write(addr, val, 8);
+address_space_rw(as, addr, (void *) val, 8, 1);
}
-void stq_le_phys(hwaddr addr, uint64_t val)
+void stq_le_phys(AddressSpace *as, hwaddr addr, uint64_t val)
{
val = cpu_to_le64(val);
-cpu_physical_memory_write(addr, val, 8);
+address_space_rw(as, addr, (void *) val, 8, 1);
}
-void stq_be_phys(hwaddr addr, uint64_t val)
+void stq_be_phys(AddressSpace *as, hwaddr addr, uint64_t val)
{
val = cpu_to_be64(val);
-cpu_physical_memory_write(addr, val, 8);
+address_space_rw(as, addr, (void *) val, 8, 1);
}
/* virtual memory access for debug (includes writing to ROM) */
diff --git a/hw/alpha/dp264.c b/hw/alpha/dp264.c
index 20795ac..1351ba5 100644
--- a/hw/alpha/dp264.c
+++ b/hw/alpha/dp264.c
@@ -161,8 +161,9 @@ static void clipper_init(QEMUMachineInitArgs *args)
load_image_targphys(initrd_filename, initrd_base,
ram_size - initrd_base);
-stq_phys(param_offset + 0x100, initrd_base +
0xfc00ULL);
-stq_phys(param_offset + 0x108, initrd_size);
+stq_phys(address_space_memory,
+ param_offset + 0x100, initrd_base +
0xfc00ULL);
+stq_phys(address_space_memory, param_offset + 0x108, initrd_size);
}
}
}
diff --git a/hw/net/vmware_utils.h b/hw/net/vmware_utils.h
index 7877cb6..4cf0e79 100644
--- a/hw/net/vmware_utils.h
+++ b/hw/net/vmware_utils.h
@@ -119,7 +119,7 @@ static inline void
vmw_shmem_st64(hwaddr addr, uint64_t value)
{
VMW_SHPRN(SHMEM store64: % PRIx64 (value % PRIx64 ), addr, value);
-stq_le_phys(addr, value);
+stq_le_phys(address_space_memory, addr, value);
}
/* Macros for simplification of operations on array-style registers */
diff --git a/hw/ppc/spapr_hcall.c b/hw/ppc/spapr_hcall.c
index a2893ad..5ce43ab 100644
--- a/hw/ppc/spapr_hcall.c
+++ b/hw/ppc/spapr_hcall.c
@@ -559,6 +559,8 @@ static target_ulong h_logical_load(PowerPCCPU *cpu,
sPAPREnvironment *spapr,
static target_ulong h_logical_store(PowerPCCPU *cpu, sPAPREnvironment *spapr,
target_ulong opcode, target_ulong *args)
{
+CPUState *cs = CPU(cpu);
+
target_ulong size = args[0];
target_ulong addr = args[1];
target_ulong val = args[2];
@@ -574,7 +576,7 @@ static target_ulong h_logical_store(PowerPCCPU *cpu,
sPAPREnvironment *spapr,
stl_phys(addr, val);
return H_SUCCESS;
case 8:
-stq_phys(addr, val);
+stq_phys(cs-as, addr, val);
return H_SUCCESS;
}
return H_PARAMETER;
@@ -639,7 +641,7 @@ static target_ulong h_logical_memop(PowerPCCPU *cpu,
sPAPREnvironment *spapr,
stl_phys(dst, tmp);
break;
case 3:
-stq_phys(dst, tmp);
+stq_phys(cs-as, dst, tmp);
break;
}
dst = dst + step;
diff --git a/hw/s390x/s390-virtio-bus.c b/hw/s390x/s390-virtio-bus.c
index 83e7287..87a1591 100644
--- a/hw/s390x/s390-virtio-bus.c
+++ b/hw/s390x/s390-virtio-bus.c
@@ -378,7 +378,8 @@ void s390_virtio_device_sync(VirtIOS390Device *dev)
vring = s390_virtio_next_ring(bus);
virtio_queue_set_addr(dev-vdev, i, vring);
virtio_queue_set_vector(dev-vdev, i, i);
-stq_be_phys(vq + VIRTIO_VQCONFIG_OFFS_ADDRESS, vring);
+stq_be_phys(address_space_memory,
+vq + VIRTIO_VQCONFIG_OFFS_ADDRESS, vring);
stw_be_phys(vq + VIRTIO_VQCONFIG_OFFS_NUM,
virtio_queue_get_num(dev-vdev, i));
}
diff --git a/hw/s390x/virtio-ccw.c b/hw/s390x/virtio-ccw.c
index 50d852a..20ad77e 100644
--- a/hw/s390x/virtio-ccw.c
+++ b/hw/s390x/virtio-ccw.c
@@ -873,7 +873,7 @@ static void
Re: [Qemu-devel] [PATCH 20/42] input: mouse: add graphic_rotate support
Hi, It looks like doing it right for relative is easy: But what is the point when this isn't used anyway? In what sense it is not used? Old code is obviously broken for the relative case. Seems to be used by hw/arm/spitz.c, which probably emulates a absolute pointing device. Machine type name contains 'PDA', so I guess a touch screen ... cheers, Gerd
[Qemu-devel] [PATCH v4 09/22] exec: Make ldq/ldub_*_phys input an AddressSpace
From: Edgar E. Iglesias edgar.igles...@xilinx.com
Signed-off-by: Edgar E. Iglesias edgar.igles...@xilinx.com
---
exec.c| 22 +++
hw/alpha/typhoon.c| 2 +-
hw/display/sm501.c| 1 +
hw/display/sm501_template.h | 2 +-
hw/net/vmware_utils.h | 4 +-
hw/ppc/spapr_hcall.c | 11 ++--
hw/s390x/s390-virtio-bus.c| 5 +-
hw/s390x/virtio-ccw.c | 18 +++---
hw/scsi/megasas.c | 3 +-
hw/virtio/virtio.c| 2 +-
include/exec/cpu-common.h | 8 +--
monitor.c | 2 +-
target-alpha/helper.c | 7 ++-
target-alpha/helper.h | 2 +-
target-alpha/mem_helper.c | 11 ++--
target-alpha/translate.c | 2 +-
target-arm/helper.c | 3 +-
target-i386/arch_memory_mapping.c | 10 ++--
target-i386/helper.c | 20 +++
target-i386/smm_helper.c | 32 +--
target-i386/svm_helper.c | 117 +++---
target-ppc/mmu-hash64.h | 7 ++-
target-s390x/helper.c | 3 +-
target-s390x/mem_helper.c | 3 +-
target-sparc/ldst_helper.c| 25
25 files changed, 185 insertions(+), 137 deletions(-)
diff --git a/exec.c b/exec.c
index aa208be..3bc4497 100644
--- a/exec.c
+++ b/exec.c
@@ -1597,7 +1597,7 @@ static uint64_t watch_mem_read(void *opaque, hwaddr addr,
{
check_watchpoint(addr ~TARGET_PAGE_MASK, ~(size - 1), BP_MEM_READ);
switch (size) {
-case 1: return ldub_phys(addr);
+case 1: return ldub_phys(address_space_memory, addr);
case 2: return lduw_phys(addr);
case 4: return ldl_phys(address_space_memory, addr);
default: abort();
@@ -2403,7 +2403,7 @@ uint32_t ldl_be_phys(AddressSpace *as, hwaddr addr)
}
/* warning: addr must be aligned */
-static inline uint64_t ldq_phys_internal(hwaddr addr,
+static inline uint64_t ldq_phys_internal(AddressSpace *as, hwaddr addr,
enum device_endian endian)
{
uint8_t *ptr;
@@ -2412,7 +2412,7 @@ static inline uint64_t ldq_phys_internal(hwaddr addr,
hwaddr l = 8;
hwaddr addr1;
-mr = address_space_translate(address_space_memory, addr, addr1, l,
+mr = address_space_translate(as, addr, addr1, l,
false);
if (l 8 || !memory_access_is_direct(mr, false)) {
/* I/O case */
@@ -2446,26 +2446,26 @@ static inline uint64_t ldq_phys_internal(hwaddr addr,
return val;
}
-uint64_t ldq_phys(hwaddr addr)
+uint64_t ldq_phys(AddressSpace *as, hwaddr addr)
{
-return ldq_phys_internal(addr, DEVICE_NATIVE_ENDIAN);
+return ldq_phys_internal(as, addr, DEVICE_NATIVE_ENDIAN);
}
-uint64_t ldq_le_phys(hwaddr addr)
+uint64_t ldq_le_phys(AddressSpace *as, hwaddr addr)
{
-return ldq_phys_internal(addr, DEVICE_LITTLE_ENDIAN);
+return ldq_phys_internal(as, addr, DEVICE_LITTLE_ENDIAN);
}
-uint64_t ldq_be_phys(hwaddr addr)
+uint64_t ldq_be_phys(AddressSpace *as, hwaddr addr)
{
-return ldq_phys_internal(addr, DEVICE_BIG_ENDIAN);
+return ldq_phys_internal(as, addr, DEVICE_BIG_ENDIAN);
}
/* XXX: optimize */
-uint32_t ldub_phys(hwaddr addr)
+uint32_t ldub_phys(AddressSpace *as, hwaddr addr)
{
uint8_t val;
-cpu_physical_memory_read(addr, val, 1);
+address_space_rw(as, addr, val, 1, 0);
return val;
}
diff --git a/hw/alpha/typhoon.c b/hw/alpha/typhoon.c
index 71a5a37..67a1070 100644
--- a/hw/alpha/typhoon.c
+++ b/hw/alpha/typhoon.c
@@ -613,7 +613,7 @@ static bool make_iommu_tlbe(hwaddr taddr, hwaddr mask,
IOMMUTLBEntry *ret)
translation, given the address of the PTE. */
static bool pte_translate(hwaddr pte_addr, IOMMUTLBEntry *ret)
{
-uint64_t pte = ldq_phys(pte_addr);
+uint64_t pte = ldq_phys(address_space_memory, pte_addr);
/* Check valid bit. */
if ((pte 1) == 0) {
diff --git a/hw/display/sm501.c b/hw/display/sm501.c
index c75d6ac..0b5f993 100644
--- a/hw/display/sm501.c
+++ b/hw/display/sm501.c
@@ -30,6 +30,7 @@
#include hw/sysbus.h
#include qemu/range.h
#include ui/pixel_ops.h
+#include exec/address-spaces.h
/*
* Status: 2010/05/07
diff --git a/hw/display/sm501_template.h b/hw/display/sm501_template.h
index 2d4a3d8..d4cea9e 100644
--- a/hw/display/sm501_template.h
+++ b/hw/display/sm501_template.h
@@ -120,7 +120,7 @@ static void glue(draw_hwc_line_, PIXEL_NAME)(SM501State *
s, int crt,
/* get pixel value */
if (i % 4 == 0) {
-bitset = ldub_phys(cursor_addr);
+bitset = ldub_phys(address_space_memory, cursor_addr);
cursor_addr++;
}
v = bitset 3;
diff --git a/hw/net/vmware_utils.h b/hw/net/vmware_utils.h
index b465eb6..eb98d2a 100644
--- a/hw/net/vmware_utils.h
+++ b/hw/net/vmware_utils.h
@@ -65,7 +65,7 @@ vmw_shmem_set(hwaddr addr, uint8 val, int len)
[Qemu-devel] [PATCH v4 17/22] exec: Make cpu_memory_rw_debug use the CPUs AS
From: Edgar E. Iglesias edgar.igles...@xilinx.com
Signed-off-by: Edgar E. Iglesias edgar.igles...@xilinx.com
---
exec.c | 10 +-
1 file changed, 5 insertions(+), 5 deletions(-)
diff --git a/exec.c b/exec.c
index 7045f76..8c05087 100644
--- a/exec.c
+++ b/exec.c
@@ -2717,11 +2717,11 @@ int cpu_memory_rw_debug(CPUState *cpu, target_ulong
addr,
if (l len)
l = len;
phys_addr += (addr ~TARGET_PAGE_MASK);
-if (is_write)
-cpu_physical_memory_write_rom(address_space_memory,
- phys_addr, buf, l);
-else
-cpu_physical_memory_rw(phys_addr, buf, l, is_write);
+if (is_write) {
+cpu_physical_memory_write_rom(cpu-as, phys_addr, buf, l);
+} else {
+address_space_rw(cpu-as, phys_addr, buf, l, 0);
+}
len -= l;
buf += l;
addr += l;
--
1.8.1.2
[Qemu-devel] [PATCH v4 18/22] memory: Add address_space_find_by_name()
From: Edgar E. Iglesias edgar.igles...@xilinx.com
Signed-off-by: Edgar E. Iglesias edgar.igles...@xilinx.com
---
include/exec/memory.h | 9 +
memory.c | 12
translate-all.c | 10 --
3 files changed, 29 insertions(+), 2 deletions(-)
diff --git a/include/exec/memory.h b/include/exec/memory.h
index 9101fc3..8465d5b 100644
--- a/include/exec/memory.h
+++ b/include/exec/memory.h
@@ -1054,4 +1054,13 @@ void address_space_unmap(AddressSpace *as, void *buffer,
hwaddr len,
#endif
+/**
+ * address_space_find_by_name: Find an AddressSpace * by name
+ *
+ * Returns an AddressSpace * if found.
+ *
+ * @name: name of an address space too look for.
+ */
+AddressSpace *address_space_find_by_name(const char *name);
+
#endif
diff --git a/memory.c b/memory.c
index 59ecc28..22fbe16 100644
--- a/memory.c
+++ b/memory.c
@@ -1720,6 +1720,18 @@ void address_space_init(AddressSpace *as, MemoryRegion
*root, const char *name)
memory_region_transaction_commit();
}
+AddressSpace *address_space_find_by_name(const char *name)
+{
+AddressSpace *as;
+
+QTAILQ_FOREACH(as, address_spaces, address_spaces_link) {
+if (strcmp(as-name, name) == 0) {
+return as;
+}
+}
+return NULL;
+}
+
void address_space_destroy(AddressSpace *as)
{
/* Flush out anything from MemoryListeners listening in on this */
diff --git a/translate-all.c b/translate-all.c
index 1ac0246..9950c3b 100644
--- a/translate-all.c
+++ b/translate-all.c
@@ -52,9 +52,8 @@
#include libutil.h
#endif
#endif
-#else
-#include exec/address-spaces.h
#endif
+#include exec/address-spaces.h
#include exec/cputlb.h
#include translate-all.h
@@ -1565,6 +1564,13 @@ void cpu_interrupt(CPUState *cpu, int mask)
cpu-tcg_exit_req = 1;
}
+/* Find an address space by name in user emulation. */
+AddressSpace *address_space_find_by_name(const char *name)
+{
+/* Unsupported. */
+return NULL;
+}
+
/*
* Walks guest process memory regions one by one
* and calls callback function 'fn' for each region.
--
1.8.1.2
[Qemu-devel] [PATCH v4 13/22] exec: Make stl_phys_notdirty input an AddressSpace
From: Edgar E. Iglesias edgar.igles...@xilinx.com
Signed-off-by: Edgar E. Iglesias edgar.igles...@xilinx.com
---
exec.c| 4 ++--
hw/arm/boot.c | 5 +++--
hw/arm/highbank.c | 6 +++---
include/exec/cpu-common.h | 2 +-
target-i386/helper.c | 16
target-sparc/mmu_helper.c | 2 +-
6 files changed, 18 insertions(+), 17 deletions(-)
diff --git a/exec.c b/exec.c
index a1f720c..9d4a4e0 100644
--- a/exec.c
+++ b/exec.c
@@ -2531,14 +2531,14 @@ uint32_t lduw_be_phys(AddressSpace *as, hwaddr addr)
/* warning: addr must be aligned. The ram page is not masked as dirty
and the code inside is not invalidated. It is useful if the dirty
bits are used to track modified PTEs */
-void stl_phys_notdirty(hwaddr addr, uint32_t val)
+void stl_phys_notdirty(AddressSpace *as, hwaddr addr, uint32_t val)
{
uint8_t *ptr;
MemoryRegion *mr;
hwaddr l = 4;
hwaddr addr1;
-mr = address_space_translate(address_space_memory, addr, addr1, l,
+mr = address_space_translate(as, addr, addr1, l,
true);
if (l 4 || !memory_access_is_direct(mr, true)) {
io_mem_write(mr, addr1, val, 4);
diff --git a/hw/arm/boot.c b/hw/arm/boot.c
index 4036262..dc62918 100644
--- a/hw/arm/boot.c
+++ b/hw/arm/boot.c
@@ -16,6 +16,7 @@
#include elf.h
#include sysemu/device_tree.h
#include qemu/config-file.h
+#include exec/address-spaces.h
/* Kernel boot protocol is specified in the kernel docs
* Documentation/arm/Booting and Documentation/arm64/booting.txt
@@ -169,7 +170,7 @@ static void default_reset_secondary(ARMCPU *cpu,
{
CPUARMState *env = cpu-env;
-stl_phys_notdirty(info-smp_bootreg_addr, 0);
+stl_phys_notdirty(address_space_memory, info-smp_bootreg_addr, 0);
env-regs[15] = info-smp_loader_start;
}
@@ -179,7 +180,7 @@ static inline bool have_dtb(const struct arm_boot_info
*info)
}
#define WRITE_WORD(p, value) do { \
-stl_phys_notdirty(p, value); \
+stl_phys_notdirty(address_space_memory, p, value); \
p += 4; \
} while (0)
diff --git a/hw/arm/highbank.c b/hw/arm/highbank.c
index d76a1d1..f66d57b 100644
--- a/hw/arm/highbank.c
+++ b/hw/arm/highbank.c
@@ -69,11 +69,11 @@ static void hb_reset_secondary(ARMCPU *cpu, const struct
arm_boot_info *info)
switch (info-nb_cpus) {
case 4:
-stl_phys_notdirty(SMP_BOOT_REG + 0x30, 0);
+stl_phys_notdirty(address_space_memory, SMP_BOOT_REG + 0x30, 0);
case 3:
-stl_phys_notdirty(SMP_BOOT_REG + 0x20, 0);
+stl_phys_notdirty(address_space_memory, SMP_BOOT_REG + 0x20, 0);
case 2:
-stl_phys_notdirty(SMP_BOOT_REG + 0x10, 0);
+stl_phys_notdirty(address_space_memory, SMP_BOOT_REG + 0x10, 0);
env-regs[15] = SMP_BOOT_ADDR;
break;
default:
diff --git a/include/exec/cpu-common.h b/include/exec/cpu-common.h
index d005c98..525fb62 100644
--- a/include/exec/cpu-common.h
+++ b/include/exec/cpu-common.h
@@ -102,7 +102,7 @@ void stq_be_phys(AddressSpace *as, hwaddr addr, uint64_t
val);
uint32_t lduw_phys(AddressSpace *as, hwaddr addr);
uint32_t ldl_phys(AddressSpace *as, hwaddr addr);
uint64_t ldq_phys(AddressSpace *as, hwaddr addr);
-void stl_phys_notdirty(hwaddr addr, uint32_t val);
+void stl_phys_notdirty(AddressSpace *as, hwaddr addr, uint32_t val);
void stw_phys(hwaddr addr, uint32_t val);
void stl_phys(AddressSpace *as, hwaddr addr, uint32_t val);
void stq_phys(AddressSpace *as, hwaddr addr, uint64_t val);
diff --git a/target-i386/helper.c b/target-i386/helper.c
index 0606908..55c0457 100644
--- a/target-i386/helper.c
+++ b/target-i386/helper.c
@@ -574,7 +574,7 @@ int cpu_x86_handle_mmu_fault(CPUX86State *env, target_ulong
addr,
}
if (!(pml4e PG_ACCESSED_MASK)) {
pml4e |= PG_ACCESSED_MASK;
-stl_phys_notdirty(pml4e_addr, pml4e);
+stl_phys_notdirty(cs-as, pml4e_addr, pml4e);
}
ptep = pml4e ^ PG_NX_MASK;
pdpe_addr = ((pml4e PHYS_ADDR_MASK) + (((addr 30) 0x1ff)
3))
@@ -591,7 +591,7 @@ int cpu_x86_handle_mmu_fault(CPUX86State *env, target_ulong
addr,
ptep = pdpe ^ PG_NX_MASK;
if (!(pdpe PG_ACCESSED_MASK)) {
pdpe |= PG_ACCESSED_MASK;
-stl_phys_notdirty(pdpe_addr, pdpe);
+stl_phys_notdirty(cs-as, pdpe_addr, pdpe);
}
} else
#endif
@@ -661,7 +661,7 @@ int cpu_x86_handle_mmu_fault(CPUX86State *env, target_ulong
addr,
pde |= PG_ACCESSED_MASK;
if (is_dirty)
pde |= PG_DIRTY_MASK;
-stl_phys_notdirty(pde_addr, pde);
+stl_phys_notdirty(cs-as, pde_addr, pde);
}
/* align to page_size */
pte = pde ((PHYS_ADDR_MASK ~(page_size - 1)) | 0xfff);
@@ -670,7 +670,7 @@ int
[Qemu-devel] [PATCH v4 20/22] cpu: Add address-space property
From: Edgar E. Iglesias edgar.igles...@xilinx.com
Signed-off-by: Edgar E. Iglesias edgar.igles...@xilinx.com
---
qom/cpu.c | 7 +++
1 file changed, 7 insertions(+)
diff --git a/qom/cpu.c b/qom/cpu.c
index 9d62479..fc6dbff 100644
--- a/qom/cpu.c
+++ b/qom/cpu.c
@@ -24,6 +24,7 @@
#include qemu/notify.h
#include qemu/log.h
#include sysemu/sysemu.h
+#include hw/qdev-properties.h
bool cpu_exists(int64_t id)
{
@@ -237,6 +238,11 @@ static int64_t cpu_common_get_arch_id(CPUState *cpu)
return cpu-cpu_index;
}
+static Property cpu_properties[] = {
+DEFINE_PROP_ADDRESS_SPACE(address-space, CPUState, as),
+DEFINE_PROP_END_OF_LIST(),
+};
+
static void cpu_class_init(ObjectClass *klass, void *data)
{
DeviceClass *dc = DEVICE_CLASS(klass);
@@ -253,6 +259,7 @@ static void cpu_class_init(ObjectClass *klass, void *data)
k-write_elf64_note = cpu_common_write_elf64_note;
k-gdb_read_register = cpu_common_gdb_read_register;
k-gdb_write_register = cpu_common_gdb_write_register;
+dc-props = cpu_properties;
dc-realize = cpu_common_realizefn;
/*
* Reason: CPUs still need special care by board code: wiring up
--
1.8.1.2
[Qemu-devel] [PATCH v4 21/22] petalogix-ml605: Create the CPU with object_new()
From: Edgar E. Iglesias edgar.igles...@xilinx.com
This is to allow future patches to set properties before cpu::realize().
Reviewed-by: Andreas Färber afaer...@suse.de
Signed-off-by: Edgar E. Iglesias edgar.igles...@xilinx.com
---
hw/microblaze/petalogix_ml605_mmu.c | 7 ++-
1 file changed, 2 insertions(+), 5 deletions(-)
diff --git a/hw/microblaze/petalogix_ml605_mmu.c
b/hw/microblaze/petalogix_ml605_mmu.c
index 1a87756..37cbbfd 100644
--- a/hw/microblaze/petalogix_ml605_mmu.c
+++ b/hw/microblaze/petalogix_ml605_mmu.c
@@ -75,7 +75,6 @@ static void
petalogix_ml605_init(QEMUMachineInitArgs *args)
{
ram_addr_t ram_size = args-ram_size;
-const char *cpu_model = args-cpu_model;
MemoryRegion *address_space_mem = get_system_memory();
DeviceState *dev, *dma, *eth0;
Object *ds, *cs;
@@ -89,10 +88,8 @@ petalogix_ml605_init(QEMUMachineInitArgs *args)
qemu_irq irq[32];
/* init CPUs */
-if (cpu_model == NULL) {
-cpu_model = microblaze;
-}
-cpu = cpu_mb_init(cpu_model);
+cpu = MICROBLAZE_CPU(object_new(TYPE_MICROBLAZE_CPU));
+object_property_set_bool(OBJECT(cpu), true, realized, error_abort);
/* Attach emulated BRAM through the LMB. */
memory_region_init_ram(phys_lmb_bram, NULL, petalogix_ml605.lmb_bram,
--
1.8.1.2
[Qemu-devel] [PATCH v4 14/22] exec: Make stw_*_phys input an AddressSpace
From: Edgar E. Iglesias edgar.igles...@xilinx.com
Signed-off-by: Edgar E. Iglesias edgar.igles...@xilinx.com
---
exec.c | 20 ++--
hw/net/vmware_utils.h | 2 +-
hw/ppc/ppc405_uc.c | 2 +-
hw/ppc/spapr_hcall.c | 4 ++--
hw/s390x/css.c | 3 ++-
hw/s390x/s390-virtio-bus.c | 8 +---
hw/s390x/virtio-ccw.c | 3 ++-
hw/sh4/r2d.c | 2 +-
hw/virtio/virtio.c | 10 ++
include/exec/cpu-common.h | 6 +++---
target-i386/smm_helper.c | 12 ++--
target-i386/svm_helper.c | 4 ++--
target-s390x/mem_helper.c | 3 ++-
target-sparc/ldst_helper.c | 6 +++---
14 files changed, 46 insertions(+), 39 deletions(-)
diff --git a/exec.c b/exec.c
index 9d4a4e0..3389b90 100644
--- a/exec.c
+++ b/exec.c
@@ -1613,7 +1613,7 @@ static void watch_mem_write(void *opaque, hwaddr addr,
stb_phys(addr, val);
break;
case 2:
-stw_phys(addr, val);
+stw_phys(address_space_memory, addr, val);
break;
case 4:
stl_phys(address_space_memory, addr, val);
@@ -2625,7 +2625,8 @@ void stb_phys(hwaddr addr, uint32_t val)
}
/* warning: addr must be aligned */
-static inline void stw_phys_internal(hwaddr addr, uint32_t val,
+static inline void stw_phys_internal(AddressSpace *as,
+ hwaddr addr, uint32_t val,
enum device_endian endian)
{
uint8_t *ptr;
@@ -2633,8 +2634,7 @@ static inline void stw_phys_internal(hwaddr addr,
uint32_t val,
hwaddr l = 2;
hwaddr addr1;
-mr = address_space_translate(address_space_memory, addr, addr1, l,
- true);
+mr = address_space_translate(as, addr, addr1, l, true);
if (l 2 || !memory_access_is_direct(mr, true)) {
#if defined(TARGET_WORDS_BIGENDIAN)
if (endian == DEVICE_LITTLE_ENDIAN) {
@@ -2665,19 +2665,19 @@ static inline void stw_phys_internal(hwaddr addr,
uint32_t val,
}
}
-void stw_phys(hwaddr addr, uint32_t val)
+void stw_phys(AddressSpace *as, hwaddr addr, uint32_t val)
{
-stw_phys_internal(addr, val, DEVICE_NATIVE_ENDIAN);
+stw_phys_internal(as, addr, val, DEVICE_NATIVE_ENDIAN);
}
-void stw_le_phys(hwaddr addr, uint32_t val)
+void stw_le_phys(AddressSpace *as, hwaddr addr, uint32_t val)
{
-stw_phys_internal(addr, val, DEVICE_LITTLE_ENDIAN);
+stw_phys_internal(as, addr, val, DEVICE_LITTLE_ENDIAN);
}
-void stw_be_phys(hwaddr addr, uint32_t val)
+void stw_be_phys(AddressSpace *as, hwaddr addr, uint32_t val)
{
-stw_phys_internal(addr, val, DEVICE_BIG_ENDIAN);
+stw_phys_internal(as, addr, val, DEVICE_BIG_ENDIAN);
}
/* XXX: optimize */
diff --git a/hw/net/vmware_utils.h b/hw/net/vmware_utils.h
index 2ed73af..d8f734f 100644
--- a/hw/net/vmware_utils.h
+++ b/hw/net/vmware_utils.h
@@ -89,7 +89,7 @@ static inline void
vmw_shmem_st16(hwaddr addr, uint16_t value)
{
VMW_SHPRN(SHMEM store16: % PRIx64 (value 0x%X), addr, value);
-stw_le_phys(addr, value);
+stw_le_phys(address_space_memory, addr, value);
}
static inline uint32_t
diff --git a/hw/ppc/ppc405_uc.c b/hw/ppc/ppc405_uc.c
index 47a4242..b0a59c3 100644
--- a/hw/ppc/ppc405_uc.c
+++ b/hw/ppc/ppc405_uc.c
@@ -65,7 +65,7 @@ ram_addr_t ppc405_set_bootinfo (CPUPPCState *env,
ppc4xx_bd_info_t *bd,
for (i = 0; i 6; i++) {
stb_phys(bdloc + 0x24 + i, bd-bi_enetaddr[i]);
}
-stw_be_phys(bdloc + 0x2A, bd-bi_ethspeed);
+stw_be_phys(cs-as, bdloc + 0x2A, bd-bi_ethspeed);
stl_be_phys(cs-as, bdloc + 0x2C, bd-bi_intfreq);
stl_be_phys(cs-as, bdloc + 0x30, bd-bi_busfreq);
stl_be_phys(cs-as, bdloc + 0x34, bd-bi_baudrate);
diff --git a/hw/ppc/spapr_hcall.c b/hw/ppc/spapr_hcall.c
index f47c3ec..ebf09e9 100644
--- a/hw/ppc/spapr_hcall.c
+++ b/hw/ppc/spapr_hcall.c
@@ -570,7 +570,7 @@ static target_ulong h_logical_store(PowerPCCPU *cpu,
sPAPREnvironment *spapr,
stb_phys(addr, val);
return H_SUCCESS;
case 2:
-stw_phys(addr, val);
+stw_phys(cs-as, addr, val);
return H_SUCCESS;
case 4:
stl_phys(cs-as, addr, val);
@@ -635,7 +635,7 @@ static target_ulong h_logical_memop(PowerPCCPU *cpu,
sPAPREnvironment *spapr,
stb_phys(dst, tmp);
break;
case 1:
-stw_phys(dst, tmp);
+stw_phys(cs-as, dst, tmp);
break;
case 2:
stl_phys(cs-as, dst, tmp);
diff --git a/hw/s390x/css.c b/hw/s390x/css.c
index cfa8a9b..75b04b4 100644
--- a/hw/s390x/css.c
+++ b/hw/s390x/css.c
@@ -680,7 +680,8 @@ static void css_update_chnmon(SubchDev *sch)
count = lduw_phys(address_space_memory,
channel_subsys-chnmon_area + offset);
count++;
-stw_phys(channel_subsys-chnmon_area + offset, count);
+stw_phys(address_space_memory,
+ channel_subsys-chnmon_area + offset, count);
}
[Qemu-devel] [PATCH v4 15/22] exec: Make stb_phys input an AddressSpace
From: Edgar E. Iglesias edgar.igles...@xilinx.com
Signed-off-by: Edgar E. Iglesias edgar.igles...@xilinx.com
---
exec.c | 6 +++---
hw/net/vmware_utils.h | 2 +-
hw/ppc/ppc405_uc.c | 10 +-
hw/ppc/spapr_hcall.c | 6 +++---
hw/s390x/s390-virtio-bus.c | 16 +++-
hw/s390x/s390-virtio.c | 2 +-
hw/scsi/megasas.c | 6 --
include/exec/cpu-common.h | 2 +-
target-i386/svm_helper.c | 2 +-
target-s390x/mem_helper.c | 2 +-
target-sparc/ldst_helper.c | 6 +++---
11 files changed, 34 insertions(+), 26 deletions(-)
diff --git a/exec.c b/exec.c
index 3389b90..248b824 100644
--- a/exec.c
+++ b/exec.c
@@ -1610,7 +1610,7 @@ static void watch_mem_write(void *opaque, hwaddr addr,
check_watchpoint(addr ~TARGET_PAGE_MASK, ~(size - 1), BP_MEM_WRITE);
switch (size) {
case 1:
-stb_phys(addr, val);
+stb_phys(address_space_memory, addr, val);
break;
case 2:
stw_phys(address_space_memory, addr, val);
@@ -2618,10 +2618,10 @@ void stl_be_phys(AddressSpace *as, hwaddr addr,
uint32_t val)
}
/* XXX: optimize */
-void stb_phys(hwaddr addr, uint32_t val)
+void stb_phys(AddressSpace *as, hwaddr addr, uint32_t val)
{
uint8_t v = val;
-cpu_physical_memory_write(addr, v, 1);
+address_space_rw(as, addr, v, 1, 1);
}
/* warning: addr must be aligned */
diff --git a/hw/net/vmware_utils.h b/hw/net/vmware_utils.h
index d8f734f..1099df6 100644
--- a/hw/net/vmware_utils.h
+++ b/hw/net/vmware_utils.h
@@ -74,7 +74,7 @@ static inline void
vmw_shmem_st8(hwaddr addr, uint8_t value)
{
VMW_SHPRN(SHMEM store8: % PRIx64 (value 0x%X), addr, value);
-stb_phys(addr, value);
+stb_phys(address_space_memory, addr, value);
}
static inline uint32_t
diff --git a/hw/ppc/ppc405_uc.c b/hw/ppc/ppc405_uc.c
index b0a59c3..ca520e8 100644
--- a/hw/ppc/ppc405_uc.c
+++ b/hw/ppc/ppc405_uc.c
@@ -63,27 +63,27 @@ ram_addr_t ppc405_set_bootinfo (CPUPPCState *env,
ppc4xx_bd_info_t *bd,
stl_be_phys(cs-as, bdloc + 0x1C, bd-bi_bootflags);
stl_be_phys(cs-as, bdloc + 0x20, bd-bi_ipaddr);
for (i = 0; i 6; i++) {
-stb_phys(bdloc + 0x24 + i, bd-bi_enetaddr[i]);
+stb_phys(cs-as, bdloc + 0x24 + i, bd-bi_enetaddr[i]);
}
stw_be_phys(cs-as, bdloc + 0x2A, bd-bi_ethspeed);
stl_be_phys(cs-as, bdloc + 0x2C, bd-bi_intfreq);
stl_be_phys(cs-as, bdloc + 0x30, bd-bi_busfreq);
stl_be_phys(cs-as, bdloc + 0x34, bd-bi_baudrate);
for (i = 0; i 4; i++) {
-stb_phys(bdloc + 0x38 + i, bd-bi_s_version[i]);
+stb_phys(cs-as, bdloc + 0x38 + i, bd-bi_s_version[i]);
}
for (i = 0; i 32; i++) {
-stb_phys(bdloc + 0x3C + i, bd-bi_r_version[i]);
+stb_phys(cs-as, bdloc + 0x3C + i, bd-bi_r_version[i]);
}
stl_be_phys(cs-as, bdloc + 0x5C, bd-bi_plb_busfreq);
stl_be_phys(cs-as, bdloc + 0x60, bd-bi_pci_busfreq);
for (i = 0; i 6; i++) {
-stb_phys(bdloc + 0x64 + i, bd-bi_pci_enetaddr[i]);
+stb_phys(cs-as, bdloc + 0x64 + i, bd-bi_pci_enetaddr[i]);
}
n = 0x6A;
if (flags 0x0001) {
for (i = 0; i 6; i++)
-stb_phys(bdloc + n++, bd-bi_pci_enetaddr2[i]);
+stb_phys(cs-as, bdloc + n++, bd-bi_pci_enetaddr2[i]);
}
stl_be_phys(cs-as, bdloc + n, bd-bi_opbfreq);
n += 4;
diff --git a/hw/ppc/spapr_hcall.c b/hw/ppc/spapr_hcall.c
index ebf09e9..3ffcc65 100644
--- a/hw/ppc/spapr_hcall.c
+++ b/hw/ppc/spapr_hcall.c
@@ -370,7 +370,7 @@ static target_ulong register_vpa(CPUPPCState *env,
target_ulong vpa)
tmp = ldub_phys(cs-as, env-vpa_addr + VPA_SHARED_PROC_OFFSET);
tmp |= VPA_SHARED_PROC_VAL;
-stb_phys(env-vpa_addr + VPA_SHARED_PROC_OFFSET, tmp);
+stb_phys(cs-as, env-vpa_addr + VPA_SHARED_PROC_OFFSET, tmp);
return H_SUCCESS;
}
@@ -567,7 +567,7 @@ static target_ulong h_logical_store(PowerPCCPU *cpu,
sPAPREnvironment *spapr,
switch (size) {
case 1:
-stb_phys(addr, val);
+stb_phys(cs-as, addr, val);
return H_SUCCESS;
case 2:
stw_phys(cs-as, addr, val);
@@ -632,7 +632,7 @@ static target_ulong h_logical_memop(PowerPCCPU *cpu,
sPAPREnvironment *spapr,
}
switch (esize) {
case 0:
-stb_phys(dst, tmp);
+stb_phys(cs-as, dst, tmp);
break;
case 1:
stw_phys(cs-as, dst, tmp);
diff --git a/hw/s390x/s390-virtio-bus.c b/hw/s390x/s390-virtio-bus.c
index 68f3e5a..e4fc353 100644
--- a/hw/s390x/s390-virtio-bus.c
+++ b/hw/s390x/s390-virtio-bus.c
@@ -359,15 +359,21 @@ void s390_virtio_device_sync(VirtIOS390Device *dev)
virtio_reset(dev-vdev);
/* Sync dev space */
-stb_phys(dev-dev_offs + VIRTIO_DEV_OFFS_TYPE, dev-vdev-device_id);
+stb_phys(address_space_memory,
+ dev-dev_offs + VIRTIO_DEV_OFFS_TYPE, dev-vdev-device_id);
-stb_phys(dev-dev_offs + VIRTIO_DEV_OFFS_NUM_VQ,
[Qemu-devel] [PATCH] hw/9pfs/virtio-9p-local.c: use snprintf() instead of sprintf()
We can not assume 'path' + 'ctx-fs_root' must be less than MAX_PATH,
so need use snprintf() instead of sprintf().
And also recommend to use ARRAY_SIZE instead of hard code macro for an
array size in snprintf().
Signed-off-by: Chen Gang gang.chen.5...@gmail.com
---
hw/9pfs/virtio-9p-local.c | 9 +
1 file changed, 5 insertions(+), 4 deletions(-)
diff --git a/hw/9pfs/virtio-9p-local.c b/hw/9pfs/virtio-9p-local.c
index fc93e9e..44a0380 100644
--- a/hw/9pfs/virtio-9p-local.c
+++ b/hw/9pfs/virtio-9p-local.c
@@ -159,7 +159,7 @@ static int local_create_mapped_attr_dir(FsContext *ctx,
const char *path)
char attr_dir[PATH_MAX];
char *tmp_path = g_strdup(path);
-snprintf(attr_dir, PATH_MAX, %s/%s/%s,
+snprintf(attr_dir, ARRAY_SIZE(attr_dir), %s/%s/%s,
ctx-fs_root, dirname(tmp_path), VIRTFS_META_DIR);
err = mkdir(attr_dir, 0700);
@@ -898,7 +898,8 @@ static int local_remove(FsContext *ctx, const char *path)
* directory
*/
if (S_ISDIR(stbuf.st_mode)) {
-sprintf(buffer, %s/%s/%s, ctx-fs_root, path, VIRTFS_META_DIR);
+snprintf(buffer, ARRAY_SIZE(buffer), %s/%s/%s,
+ ctx-fs_root, path, VIRTFS_META_DIR);
err = remove(buffer);
if (err 0 errno != ENOENT) {
/*
@@ -1033,8 +1034,8 @@ static int local_unlinkat(FsContext *ctx, V9fsPath *dir,
* If directory remove .virtfs_metadata contained in the
* directory
*/
-sprintf(buffer, %s/%s/%s, ctx-fs_root,
-fullname.data, VIRTFS_META_DIR);
+snprintf(buffer, ARRAY_SIZE(buffer), %s/%s/%s, ctx-fs_root,
+ fullname.data, VIRTFS_META_DIR);
ret = remove(buffer);
if (ret 0 errno != ENOENT) {
/*
--
1.7.11.7
[Qemu-devel] [PATCH v4 19/22] qdev: Add qdev property type for AddressSpaces
From: Edgar E. Iglesias edgar.igles...@xilinx.com
Signed-off-by: Edgar E. Iglesias edgar.igles...@xilinx.com
---
hw/core/qdev-properties-system.c | 8 ++
hw/core/qdev-properties.c| 54
include/hw/qdev-properties.h | 5
stubs/Makefile.objs | 1 +
stubs/memory.c | 6 +
5 files changed, 74 insertions(+)
create mode 100644 stubs/memory.c
diff --git a/hw/core/qdev-properties-system.c b/hw/core/qdev-properties-system.c
index 3f29b49..27c2899 100644
--- a/hw/core/qdev-properties-system.c
+++ b/hw/core/qdev-properties-system.c
@@ -365,6 +365,14 @@ void qdev_prop_set_netdev(DeviceState *dev, const char
*name,
value ? value-name : , name, error_abort);
}
+void qdev_prop_set_address_space(DeviceState *dev, const char *name,
+ AddressSpace *value)
+{
+assert(!value || value-name);
+object_property_set_str(OBJECT(dev),
+value ? value-name : , name, error_abort);
+}
+
void qdev_set_nic_properties(DeviceState *dev, NICInfo *nd)
{
qdev_prop_set_macaddr(dev, mac, nd-macaddr.a);
diff --git a/hw/core/qdev-properties.c b/hw/core/qdev-properties.c
index b949f0e..68e09e2 100644
--- a/hw/core/qdev-properties.c
+++ b/hw/core/qdev-properties.c
@@ -1176,3 +1176,57 @@ PropertyInfo qdev_prop_size = {
.get = get_size,
.set = set_size,
};
+
+/* --- AddressSpace --- */
+
+static int parse_address_space(DeviceState *dev, const char *name,
+ AddressSpace **ptr)
+{
+AddressSpace *as = address_space_find_by_name(name);
+if (as == NULL) {
+return -ENOENT;
+}
+*ptr = as;
+return 0;
+}
+
+static void get_address_space(Object *obj, Visitor *v, void *opaque,
+ const char *name, Error **errp)
+{
+AddressSpace *as = qdev_get_prop_ptr(DEVICE(obj), opaque);
+char *p = (char *) (as ? as-name : );
+
+visit_type_str(v, p, name, errp);
+}
+
+static void set_address_space(Object *obj, Visitor *v, void *opaque,
+ const char *name, Error **errp)
+{
+DeviceState *dev = DEVICE(obj);
+Property *prop = opaque;
+Error *local_err = NULL;
+AddressSpace **as = qdev_get_prop_ptr(dev, prop);
+char *str;
+int ret;
+
+if (dev-realized) {
+qdev_prop_set_after_realize(dev, name, errp);
+return;
+}
+
+visit_type_str(v, str, name, local_err);
+if (local_err) {
+error_propagate(errp, local_err);
+return;
+}
+
+ret = parse_address_space(dev, str, as);
+error_set_from_qdev_prop_error(errp, ret, dev, prop, str);
+g_free(str);
+}
+
+PropertyInfo qdev_prop_address_space = {
+.name = address_space,
+.get = get_address_space,
+.set = set_address_space,
+};
diff --git a/include/hw/qdev-properties.h b/include/hw/qdev-properties.h
index 77c6f7c..020439f 100644
--- a/include/hw/qdev-properties.h
+++ b/include/hw/qdev-properties.h
@@ -19,6 +19,7 @@ extern PropertyInfo qdev_prop_size;
extern PropertyInfo qdev_prop_string;
extern PropertyInfo qdev_prop_chr;
extern PropertyInfo qdev_prop_ptr;
+extern PropertyInfo qdev_prop_address_space;
extern PropertyInfo qdev_prop_macaddr;
extern PropertyInfo qdev_prop_losttickpolicy;
extern PropertyInfo qdev_prop_bios_chs_trans;
@@ -141,6 +142,8 @@ extern PropertyInfo qdev_prop_arraylen;
#define DEFINE_PROP_PTR(_n, _s, _f) \
DEFINE_PROP(_n, _s, _f, qdev_prop_ptr, void*)
+#define DEFINE_PROP_ADDRESS_SPACE(_n, _s, _f) \
+DEFINE_PROP(_n, _s, _f, qdev_prop_address_space, AddressSpace*)
#define DEFINE_PROP_CHR(_n, _s, _f) \
DEFINE_PROP(_n, _s, _f, qdev_prop_chr, CharDriverState*)
#define DEFINE_PROP_STRING(_n, _s, _f) \
@@ -177,6 +180,8 @@ void qdev_prop_set_uint32(DeviceState *dev, const char
*name, uint32_t value);
void qdev_prop_set_int32(DeviceState *dev, const char *name, int32_t value);
void qdev_prop_set_uint64(DeviceState *dev, const char *name, uint64_t value);
void qdev_prop_set_string(DeviceState *dev, const char *name, const char
*value);
+void qdev_prop_set_address_space(DeviceState *dev, const char *name,
+ AddressSpace *value);
void qdev_prop_set_chr(DeviceState *dev, const char *name, CharDriverState
*value);
void qdev_prop_set_netdev(DeviceState *dev, const char *name, NetClientState
*value);
int qdev_prop_set_drive(DeviceState *dev, const char *name, BlockDriverState
*value) QEMU_WARN_UNUSED_RESULT;
diff --git a/stubs/Makefile.objs b/stubs/Makefile.objs
index df92fe5..9e1e5da 100644
--- a/stubs/Makefile.objs
+++ b/stubs/Makefile.objs
@@ -26,4 +26,5 @@ stub-obj-y += uuid.o
stub-obj-y += vm-stop.o
stub-obj-y += vmstate.o
stub-obj-$(CONFIG_WIN32) += fd-register.o
+stub-obj-y += memory.o
stub-obj-y += cpus.o
diff --git a/stubs/memory.c b/stubs/memory.c
new file mode 100644
index
[Qemu-devel] [PATCH v4 22/22] petalogix-ml605: Make the LMB visible only to the CPU
From: Edgar E. Iglesias edgar.igles...@xilinx.com Signed-off-by: Edgar E. Iglesias edgar.igles...@xilinx.com --- hw/microblaze/petalogix_ml605_mmu.c | 17 - 1 file changed, 16 insertions(+), 1 deletion(-) diff --git a/hw/microblaze/petalogix_ml605_mmu.c b/hw/microblaze/petalogix_ml605_mmu.c index 37cbbfd..d1acbff 100644 --- a/hw/microblaze/petalogix_ml605_mmu.c +++ b/hw/microblaze/petalogix_ml605_mmu.c @@ -86,16 +86,31 @@ petalogix_ml605_init(QEMUMachineInitArgs *args) MemoryRegion *phys_lmb_bram = g_new(MemoryRegion, 1); MemoryRegion *phys_ram = g_new(MemoryRegion, 1); qemu_irq irq[32]; +MemoryRegion *sysmem_alias = g_new(MemoryRegion, 1); +MemoryRegion *mr_cpu_root = g_new(MemoryRegion, 1); +AddressSpace *as_cpu = g_malloc0(sizeof(*as_cpu)); + +/* Setup the CPU specific address-space. */ +memory_region_init(mr_cpu_root, NULL, as-cpu-root, INT64_MAX); +address_space_init(as_cpu, mr_cpu_root, as/cpu); /* init CPUs */ cpu = MICROBLAZE_CPU(object_new(TYPE_MICROBLAZE_CPU)); +qdev_prop_set_address_space(DEVICE(cpu), address-space, as_cpu); object_property_set_bool(OBJECT(cpu), true, realized, error_abort); +/* Populate the CPU AS with the LMB only visible to the CPU. */ +memory_region_init_alias(sysmem_alias, NULL, sysmem_alias, + address_space_mem, 0, + memory_region_size(address_space_mem)); +memory_region_add_subregion(mr_cpu_root, 0x, sysmem_alias); + /* Attach emulated BRAM through the LMB. */ memory_region_init_ram(phys_lmb_bram, NULL, petalogix_ml605.lmb_bram, LMB_BRAM_SIZE); vmstate_register_ram_global(phys_lmb_bram); -memory_region_add_subregion(address_space_mem, 0x, phys_lmb_bram); +memory_region_add_subregion_overlap(mr_cpu_root, 0x, +phys_lmb_bram, 2); memory_region_init_ram(phys_ram, NULL, petalogix_ml605.ram, ram_size); vmstate_register_ram_global(phys_ram); -- 1.8.1.2
Re: [Qemu-devel] [PATCH 03/10] block: Make bdrv_file_open() static
Am 31.01.2014 um 21:20 hat Max Reitz geschrieben:
On 29.01.2014 14:26, Kevin Wolf wrote:
Am 26.01.2014 um 20:02 hat Max Reitz geschrieben:
Add the bdrv_open() option BDRV_O_PROTOCOL which results in passing the
call to bdrv_file_open(). Additionally, make bdrv_file_open() static and
therefore bdrv_open() the only way to call it.
Consequently, all existing calls to bdrv_file_open() have to be adjusted
to use bdrv_open() with the BDRV_O_PROTOCOL flag instead.
Signed-off-by: Max Reitz mre...@redhat.com
---
block.c | 17 -
block/cow.c | 6 +++---
block/qcow.c | 6 +++---
block/qcow2.c | 5 +++--
block/qed.c | 5 +++--
block/sheepdog.c | 8 +---
block/vhdx.c | 5 +++--
block/vmdk.c | 11 +++
include/block/block.h | 5 ++---
qemu-io.c | 4 +++-
10 files changed, 44 insertions(+), 28 deletions(-)
diff --git a/include/block/block.h b/include/block/block.h
index a421041..396f9ed 100644
--- a/include/block/block.h
+++ b/include/block/block.h
@@ -102,6 +102,8 @@ typedef enum {
#define BDRV_O_CHECK 0x1000 /* open solely for consistency check */
#define BDRV_O_ALLOW_RDWR 0x2000 /* allow reopen to change from r/o to
r/w */
#define BDRV_O_UNMAP 0x4000 /* execute guest UNMAP/TRIM operations
*/
+#define BDRV_O_PROTOCOL0x8000 /* open the file using a protocol
instead of
+ a block driver */
Protocol drivers are a subset of block drivers, so this description
doesn't make sense.
Hm, technically they probably are but it always seemed to me that
bdrv_open() would never directly use a protocol, instead using raw
as the format if no format was found.
Except if you explicitly specify something like format=file.
More importantly, it is actually possible to use a non-protocol
block driver with BDRV_O_PROTOCOL; it just needs to be explicitly
specified.
Yes, I think with explicit specification of the driver you get mostly
the same results with bdrv_open() and bdrv_file_open().
I guess we need to list the differences between bdrv_open() and
bdrv_file_open() in order to define what this flag really changes. I
think this includes:
- Disables format probing
- BDRV_O_SNAPSHOT is ignored
- No backing files are opened
- Probably a few more
So, to me, the main difference is that bdrv_open() always uses some
non-protocol block driver, whereas bdrv_file_open() only probes for
protocol block drivers (if a non-protocol driver should be used, it
has to be explicitly specified).
The current comment for bdrv_file_open() doesn't help much, either:
“Opens a file using a protocol”. Therefore, the easiest way would
just be to state “behaves like the old bdrv_file_open()”, but that
would not be very helpful. Perhaps we could formulate it like “Opens
a single file (no backing chain, etc.) using only a protocol driver
deduced from the filename, if not explicitly specified otherwise.”
Perhaps we should really specify it as the difference in probing:
- bdrv_open() adds a probed format layer on top of bs
- bdrv_file_open() probes protocols for bs
All other differences can probably go away without breaking anything:
BDRV_O_SNAPSHOT can be hopefully be moved to drive_init() (the tricky one
here is qmp_change_blockdev), and supporting backing files in
bdrv_file_open() by moving their handling to common code shouldn't hurt.
Any other differences that need to be eliminated? Once we know
what the differences should be, I guess we can greatly simplify the
implementation.
Kevin
Re: [Qemu-devel] [PATCH 0/2] qtest: don't leak pid files and UNIX domain sockets
On Fri, Jan 31, 2014 at 12:07:34AM +, Peter Maydell wrote: On 21 November 2013 11:03, Stefan Hajnoczi stefa...@redhat.com wrote: GLib uses abort(3) to exit failed test cases. As a result, the pid file and UNIX domain sockets for a running test are leaked upon failure. Since abort(3) does not call atexit(3) handler functions, we could set up a SIGABRT handler that performs cleanup. But there are other conditions where processes die, like SIGSEGV or SIGBUS. Let's unlink pid files and UNIX domain sockets as soon as the QEMU process has initialized and connections have been made. This eliminates the possibility of leaking these files. So looking back through mailing list history suggests that these patches are supposed to avoid intermittent make check failures like: TEST: tests/qom-test... (pid=5078) /i386/qom/none: ** ERROR:/home/petmay01/linaro/qemu-for-merges/tests/libqtest.c:71:init_socket: assertion failed (ret != -1): (-1 != -1) FAIL GTester: last random seed: R02S79ea313790bc9a8b21d9af5ed55c2fff (pid=5080) /i386/qom/pc:OK /i386/qom/isapc: OK /i386/qom/q35: OK FAIL: tests/qom-test but this patch series doesn't actually say that's what it's for, so does it fix that kind of error? I still think we should merge these patches :). Are you happy to merge them? Stefan
[Qemu-devel] [PATCH v4 16/22] exec: Make cpu_physical_memory_write_rom input an AS
From: Edgar E. Iglesias edgar.igles...@xilinx.com
Signed-off-by: Edgar E. Iglesias edgar.igles...@xilinx.com
---
exec.c| 15 ---
hw/core/loader.c | 3 ++-
hw/intc/apic.c| 3 ++-
hw/sparc/sun4m.c | 3 ++-
include/exec/cpu-common.h | 2 +-
5 files changed, 15 insertions(+), 11 deletions(-)
diff --git a/exec.c b/exec.c
index 248b824..7045f76 100644
--- a/exec.c
+++ b/exec.c
@@ -2099,7 +2099,7 @@ enum write_rom_type {
FLUSH_CACHE,
};
-static inline void cpu_physical_memory_write_rom_internal(
+static inline void cpu_physical_memory_write_rom_internal(AddressSpace *as,
hwaddr addr, const uint8_t *buf, int len, enum write_rom_type type)
{
hwaddr l;
@@ -2109,8 +2109,7 @@ static inline void cpu_physical_memory_write_rom_internal(
while (len 0) {
l = len;
-mr = address_space_translate(address_space_memory,
- addr, addr1, l, true);
+mr = address_space_translate(as, addr, addr1, l, true);
if (!(memory_region_is_ram(mr) ||
memory_region_is_romd(mr))) {
@@ -2136,10 +2135,10 @@ static inline void
cpu_physical_memory_write_rom_internal(
}
/* used for ROM loading : can write in RAM and ROM */
-void cpu_physical_memory_write_rom(hwaddr addr,
+void cpu_physical_memory_write_rom(AddressSpace *as, hwaddr addr,
const uint8_t *buf, int len)
{
-cpu_physical_memory_write_rom_internal(addr, buf, len, WRITE_DATA);
+cpu_physical_memory_write_rom_internal(as, addr, buf, len, WRITE_DATA);
}
void cpu_flush_icache_range(hwaddr start, int len)
@@ -2154,7 +2153,8 @@ void cpu_flush_icache_range(hwaddr start, int len)
return;
}
-cpu_physical_memory_write_rom_internal(start, NULL, len, FLUSH_CACHE);
+cpu_physical_memory_write_rom_internal(address_space_memory,
+ start, NULL, len, FLUSH_CACHE);
}
typedef struct {
@@ -2718,7 +2718,8 @@ int cpu_memory_rw_debug(CPUState *cpu, target_ulong addr,
l = len;
phys_addr += (addr ~TARGET_PAGE_MASK);
if (is_write)
-cpu_physical_memory_write_rom(phys_addr, buf, l);
+cpu_physical_memory_write_rom(address_space_memory,
+ phys_addr, buf, l);
else
cpu_physical_memory_rw(phys_addr, buf, l, is_write);
len -= l;
diff --git a/hw/core/loader.c b/hw/core/loader.c
index 0634bee..e1c3f3a 100644
--- a/hw/core/loader.c
+++ b/hw/core/loader.c
@@ -778,7 +778,8 @@ static void rom_reset(void *unused)
void *host = memory_region_get_ram_ptr(rom-mr);
memcpy(host, rom-data, rom-datasize);
} else {
-cpu_physical_memory_write_rom(rom-addr, rom-data, rom-datasize);
+cpu_physical_memory_write_rom(address_space_memory,
+ rom-addr, rom-data, rom-datasize);
}
if (rom-isrom) {
/* rom needs to be written only once */
diff --git a/hw/intc/apic.c b/hw/intc/apic.c
index 3d3deb6..361ae90 100644
--- a/hw/intc/apic.c
+++ b/hw/intc/apic.c
@@ -129,7 +129,8 @@ static void apic_sync_vapic(APICCommonState *s, int
sync_type)
}
vapic_state.irr = vector 0xff;
-cpu_physical_memory_write_rom(s-vapic_paddr + start,
+cpu_physical_memory_write_rom(address_space_memory,
+ s-vapic_paddr + start,
((void *)vapic_state) + start, length);
}
}
diff --git a/hw/sparc/sun4m.c b/hw/sparc/sun4m.c
index 94f7950..2957d90 100644
--- a/hw/sparc/sun4m.c
+++ b/hw/sparc/sun4m.c
@@ -577,7 +577,8 @@ static void idreg_init(hwaddr addr)
s = SYS_BUS_DEVICE(dev);
sysbus_mmio_map(s, 0, addr);
-cpu_physical_memory_write_rom(addr, idreg_data, sizeof(idreg_data));
+cpu_physical_memory_write_rom(address_space_memory,
+ addr, idreg_data, sizeof(idreg_data));
}
#define MACIO_ID_REGISTER(obj) \
diff --git a/include/exec/cpu-common.h b/include/exec/cpu-common.h
index d0fe123..a21b65a 100644
--- a/include/exec/cpu-common.h
+++ b/include/exec/cpu-common.h
@@ -108,7 +108,7 @@ void stl_phys(AddressSpace *as, hwaddr addr, uint32_t val);
void stq_phys(AddressSpace *as, hwaddr addr, uint64_t val);
#endif
-void cpu_physical_memory_write_rom(hwaddr addr,
+void cpu_physical_memory_write_rom(AddressSpace *as, hwaddr addr,
const uint8_t *buf, int len);
void cpu_flush_icache_range(hwaddr start, int len);
--
1.8.1.2
Re: [Qemu-devel] [PATCH] opencores_eth: flush queue whenever can_receive can go from false to true
On Mon, Feb 03, 2014 at 08:20:02AM +0400, Max Filippov wrote: The following registers control whether MAC can receive frames: - MODER.RXEN bit that enables/disables receiver; - TX_BD_NUM register that specifies number of RX descriptors. Notify QEMU networking core when the MAC is ready to receive frames. Discard frame and raise BUSY interrupt when the frame arrives but the current RX descriptor is not empty. Signed-off-by: Max Filippov jcmvb...@gmail.com --- hw/net/opencores_eth.c | 33 +++-- 1 file changed, 31 insertions(+), 2 deletions(-) Thanks, applied to my net tree: https://github.com/stefanha/qemu/commits/net Stefan
Re: [Qemu-devel] [PATCH] configure: use glib in glib pkg-config check.
On Mon, Feb 03, 2014 at 03:26:15PM +1100, Chris Johns wrote: Building against with a recent glib in a custom prefix fails because the gthread cflags in the pkg-config file do not have the correct path while the glib pc file does. Signed-off-by: Chris Johns chr...@rtems.org --- configure | 6 +++--- 1 file changed, 3 insertions(+), 3 deletions(-) diff --git a/configure b/configure index b472694..12f730f 100755 --- a/configure +++ b/configure @@ -2343,9 +2343,9 @@ if test $mingw32 = yes; then else glib_req_ver=2.12 fi -if $pkg_config --atleast-version=$glib_req_ver gthread-2.0; then -glib_cflags=`$pkg_config --cflags gthread-2.0` -glib_libs=`$pkg_config --libs gthread-2.0` +if $pkg_config --atleast-version=$glib_req_ver glib-2.0; then +glib_cflags=`$pkg_config --cflags glib-2.0` +glib_libs=`$pkg_config --libs glib-2.0` LIBS=$glib_libs $LIBS libs_qga=$glib_libs $libs_qga else This change will cause -pthread linker + compiler flag to be lost. What glib version are you seeing a problem with ? It seems we should really fix glib, since this will affect countless 1000's of apps using it, not try to workaround in all downstream apps. Daniel -- |: http://berrange.com -o-http://www.flickr.com/photos/dberrange/ :| |: http://libvirt.org -o- http://virt-manager.org :| |: http://autobuild.org -o- http://search.cpan.org/~danberr/ :| |: http://entangle-photo.org -o- http://live.gnome.org/gtk-vnc :|
Re: [Qemu-devel] [PATCH 0/2] qtest: don't leak pid files and UNIX domain sockets
On 3 February 2014 09:54, Stefan Hajnoczi stefa...@redhat.com wrote: I still think we should merge these patches :). Are you happy to merge them? They look like they're fixing a bug to me, so yes. You could make my life easier by arranging for them to appear in a pull request... thanks -- PMM
Re: [Qemu-devel] qemu-iotest 077 does not terminate with VMDK
Am 03.02.2014 um 09:44 hat Peter Lieven geschrieben: Hi, has anyone tried iotest 77 (RMW) with VMDK? for me it does not terminate. I have a patch for that (077 should be raw-only), but apparently I forgot to send it out. I'll do that now. Kevin
[Qemu-devel] [PATCH] qemu-iotest: Make 077 raw-only
The qemu-io command sequences make the assumption that an unaligned request on the format layer will be unaligned on the blkdebug layer as well. This doesn't necessarily hold true for drivers other than raw. Signed-off-by: Kevin Wolf kw...@redhat.com --- tests/qemu-iotests/077 | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/tests/qemu-iotests/077 b/tests/qemu-iotests/077 index bbf7b51..4dd1bdd 100755 --- a/tests/qemu-iotests/077 +++ b/tests/qemu-iotests/077 @@ -38,7 +38,7 @@ trap _cleanup; exit \$status 0 1 2 3 15 . ./common.rc . ./common.filter -_supported_fmt generic +_supported_fmt raw _supported_proto generic _supported_os Linux -- 1.8.1.4
Re: [Qemu-devel] [PATCH] hw/9pfs/virtio-9p-local.c: use snprintf() instead of sprintf()
On Mon, Feb 03, 2014 at 06:00:42PM +0800, Chen Gang wrote: We can not assume 'path' + 'ctx-fs_root' must be less than MAX_PATH, so need use snprintf() instead of sprintf(). And also recommend to use ARRAY_SIZE instead of hard code macro for an array size in snprintf(). In the event that there is overflow this will cause the data to be truncated, potentially causing QEMU to access the wrong file on the host. Both snprintf and sprintf are really bad because of their use of fixed buffers. Better to change it to g_strdup_printf which dynamically allocates buffers. Daniel -- |: http://berrange.com -o-http://www.flickr.com/photos/dberrange/ :| |: http://libvirt.org -o- http://virt-manager.org :| |: http://autobuild.org -o- http://search.cpan.org/~danberr/ :| |: http://entangle-photo.org -o- http://live.gnome.org/gtk-vnc :|
Re: [Qemu-devel] [PATCH] hw/9pfs/virtio-9p-local.c: use snprintf() instead of sprintf()
On 02/03/2014 06:34 PM, Daniel P. Berrange wrote: On Mon, Feb 03, 2014 at 06:00:42PM +0800, Chen Gang wrote: We can not assume 'path' + 'ctx-fs_root' must be less than MAX_PATH, so need use snprintf() instead of sprintf(). And also recommend to use ARRAY_SIZE instead of hard code macro for an array size in snprintf(). In the event that there is overflow this will cause the data to be truncated, potentially causing QEMU to access the wrong file on the host. Both snprintf and sprintf are really bad because of their use of fixed buffers. Better to change it to g_strdup_printf which dynamically allocates buffers. That sounds reasonable to me, I will send patch v2 for it. Thanks. -- Chen Gang Open, share and attitude like air, water and life which God blessed
Re: [Qemu-devel] [PATCH 6/6] linux-user: Fix trampoline code for CRIS
On Sun, Feb 02, 2014 at 03:04:52AM +, edgar.igles...@gmail.com wrote:
From: Stefan Weil s...@weilnetz.de
__put_user can write bytes, words (2 bytes) or longwords (4 bytes).
Here obviously words should have been written, but bytes were written,
so values like 0x9c5f were truncated to 0x5f.
Fix this by changing retcode from uint8_t to to uint16_t in
target_signal_frame and also in the unused rt_signal_frame.
This problem was reported by static code analysis (smatch).
Acked-by: Riku Voipio riku.voi...@linaro.org
Cc: qemu-sta...@nongnu.org
Signed-off-by: Stefan Weil s...@weilnetz.de
Reviewed-by: Peter Maydell peter.mayd...@linaro.org
Tested-by: Edgar E. Iglesias edgar.igles...@xilinx.com
Reviewed-by: Edgar E. Iglesias edgar.igles...@xilinx.com
Signed-off-by: Edgar E. Iglesias edgar.igles...@xilinx.com
---
linux-user/signal.c | 8
1 file changed, 4 insertions(+), 4 deletions(-)
diff --git a/linux-user/signal.c b/linux-user/signal.c
index 01d7c39..82e8592 100644
--- a/linux-user/signal.c
+++ b/linux-user/signal.c
@@ -3659,7 +3659,7 @@ struct target_sigcontext {
struct target_signal_frame {
struct target_sigcontext sc;
uint32_t extramask[TARGET_NSIG_WORDS - 1];
-uint8_t retcode[8]; /* Trampoline code. */
+uint16_t retcode[4]; /* Trampoline code. */
};
struct rt_signal_frame {
@@ -3667,7 +3667,7 @@ struct rt_signal_frame {
void *puc;
siginfo_t info;
struct ucontext uc;
-uint8_t retcode[8]; /* Trampoline code. */
+uint16_t retcode[4]; /* Trampoline code. */
};
static void setup_sigcontext(struct target_sigcontext *sc, CPUCRISState *env)
@@ -3745,8 +3745,8 @@ static void setup_frame(int sig, struct
target_sigaction *ka,
*/
err |= __put_user(0x9c5f, frame-retcode+0);
err |= __put_user(TARGET_NR_sigreturn,
- frame-retcode+2);
- err |= __put_user(0xe93d, frame-retcode+4);
+ frame-retcode + 1);
+ err |= __put_user(0xe93d, frame-retcode + 2);
/* Save the mask. */
err |= __put_user(set-sig[0], frame-sc.oldmask);
--
1.8.3.2
[Qemu-devel] [PATCH v4 2/5] pcihp: make PCI hotplug mmio handlers indifferent to PCI_HOTPLUG_ADDR
... removes dependency of mmio handler on PCI_HOTPLUG_ADDR.
It will be needed in case of Q35 where base could be different.
Signed-off-by: Igor Mammedov imamm...@redhat.com
---
hw/acpi/pcihp.c | 24
1 files changed, 12 insertions(+), 12 deletions(-)
diff --git a/hw/acpi/pcihp.c b/hw/acpi/pcihp.c
index 464739a..64c8cf2 100644
--- a/hw/acpi/pcihp.c
+++ b/hw/acpi/pcihp.c
@@ -48,11 +48,11 @@
#define PCI_HOTPLUG_ADDR 0xae00
#define PCI_HOTPLUG_SIZE 0x0014
-#define PCI_UP_BASE 0xae00
-#define PCI_DOWN_BASE 0xae04
-#define PCI_EJ_BASE 0xae08
-#define PCI_RMV_BASE 0xae0c
-#define PCI_SEL_BASE 0xae10
+#define PCI_UP_BASE 0x
+#define PCI_DOWN_BASE 0x0004
+#define PCI_EJ_BASE 0x0008
+#define PCI_RMV_BASE 0x000c
+#define PCI_SEL_BASE 0x0010
typedef struct AcpiPciHpFind {
int bsel;
@@ -213,24 +213,24 @@ static uint64_t pci_read(void *opaque, hwaddr addr,
unsigned int size)
}
switch (addr) {
-case PCI_UP_BASE - PCI_HOTPLUG_ADDR:
+case PCI_UP_BASE:
val = s-acpi_pcihp_pci_status[bsel].up;
s-acpi_pcihp_pci_status[bsel].up = 0;
ACPI_PCIHP_DPRINTF(pci_up_read % PRIu32 \n, val);
break;
-case PCI_DOWN_BASE - PCI_HOTPLUG_ADDR:
+case PCI_DOWN_BASE:
val = s-acpi_pcihp_pci_status[bsel].down;
ACPI_PCIHP_DPRINTF(pci_down_read % PRIu32 \n, val);
break;
-case PCI_EJ_BASE - PCI_HOTPLUG_ADDR:
+case PCI_EJ_BASE:
/* No feature defined yet */
ACPI_PCIHP_DPRINTF(pci_features_read % PRIu32 \n, val);
break;
-case PCI_RMV_BASE - PCI_HOTPLUG_ADDR:
+case PCI_RMV_BASE:
val = s-acpi_pcihp_pci_status[bsel].hotplug_enable;
ACPI_PCIHP_DPRINTF(pci_rmv_read % PRIu32 \n, val);
break;
-case PCI_SEL_BASE - PCI_HOTPLUG_ADDR:
+case PCI_SEL_BASE:
val = s-hotplug_select;
ACPI_PCIHP_DPRINTF(pci_sel_read % PRIu32 \n, val);
default:
@@ -245,7 +245,7 @@ static void pci_write(void *opaque, hwaddr addr, uint64_t
data,
{
AcpiPciHpState *s = opaque;
switch (addr) {
-case PCI_EJ_BASE - PCI_HOTPLUG_ADDR:
+case PCI_EJ_BASE:
if (s-hotplug_select = ACPI_PCIHP_MAX_HOTPLUG_BUS) {
break;
}
@@ -253,7 +253,7 @@ static void pci_write(void *opaque, hwaddr addr, uint64_t
data,
ACPI_PCIHP_DPRINTF(pciej write % HWADDR_PRIx == % PRIu64 \n,
addr, data);
break;
-case PCI_SEL_BASE - PCI_HOTPLUG_ADDR:
+case PCI_SEL_BASE:
s-hotplug_select = data;
ACPI_PCIHP_DPRINTF(pcisel write % HWADDR_PRIx == % PRIu64 \n,
addr, data);
--
1.7.1
[Qemu-devel] [PATCH v4 1/5] pcihp: replace enable|disable_device() with oneliners
enable_device() and disable_device() functions aren't reused anywere,
so replace them with respective oneliners at call sites.
Signed-off-by: Igor Mammedov imamm...@redhat.com
---
hw/acpi/pcihp.c | 14 ++
1 files changed, 2 insertions(+), 12 deletions(-)
diff --git a/hw/acpi/pcihp.c b/hw/acpi/pcihp.c
index 4345f5d..464739a 100644
--- a/hw/acpi/pcihp.c
+++ b/hw/acpi/pcihp.c
@@ -177,16 +177,6 @@ void acpi_pcihp_reset(AcpiPciHpState *s)
acpi_pcihp_update(s);
}
-static void enable_device(AcpiPciHpState *s, unsigned bsel, int slot)
-{
-s-acpi_pcihp_pci_status[bsel].up |= (1U slot);
-}
-
-static void disable_device(AcpiPciHpState *s, unsigned bsel, int slot)
-{
-s-acpi_pcihp_pci_status[bsel].down |= (1U slot);
-}
-
int acpi_pcihp_device_hotplug(AcpiPciHpState *s, PCIDevice *dev,
PCIHotplugState state)
{
@@ -204,9 +194,9 @@ int acpi_pcihp_device_hotplug(AcpiPciHpState *s, PCIDevice
*dev,
}
if (state == PCI_HOTPLUG_ENABLED) {
-enable_device(s, bsel, slot);
+s-acpi_pcihp_pci_status[bsel].up |= (1U slot);
} else {
-disable_device(s, bsel, slot);
+s-acpi_pcihp_pci_status[bsel].down |= (1U slot);
}
return 0;
--
1.7.1
[Qemu-devel] [PATCH v4 3/5] pcihp: make pci_read() mmio calback compatible with legacy ACPI hotplug
due to recent change introduced by:
pcihp: reduce number of device check events
'up' field is cleared right after it's read.
This is incompatible with legacy BIOS ACPI code
where PCNF ACPI method reads this field 32 times.
To make pci_read mmio callback compatible with legacy
'up' behavior, pcihp code will need to know in which
mode it runs add 'legacy_piix' field to AcpiPciHpState
structure and alter register behavior accordingly.
Signed-off-by: Igor Mammedov imamm...@redhat.com
---
v2:
- keep acpi-pci-hotplug-with-bridge-support property
and tweek pcihp init API to initialize .legacy_piix field
---
hw/acpi/pcihp.c |7 +--
hw/acpi/piix4.c |3 ++-
include/hw/acpi/pcihp.h |3 ++-
3 files changed, 9 insertions(+), 4 deletions(-)
diff --git a/hw/acpi/pcihp.c b/hw/acpi/pcihp.c
index 64c8cf2..974f01c 100644
--- a/hw/acpi/pcihp.c
+++ b/hw/acpi/pcihp.c
@@ -215,7 +215,9 @@ static uint64_t pci_read(void *opaque, hwaddr addr,
unsigned int size)
switch (addr) {
case PCI_UP_BASE:
val = s-acpi_pcihp_pci_status[bsel].up;
-s-acpi_pcihp_pci_status[bsel].up = 0;
+if (!s-legacy_piix) {
+s-acpi_pcihp_pci_status[bsel].up = 0;
+}
ACPI_PCIHP_DPRINTF(pci_up_read % PRIu32 \n, val);
break;
case PCI_DOWN_BASE:
@@ -273,9 +275,10 @@ static const MemoryRegionOps acpi_pcihp_io_ops = {
};
void acpi_pcihp_init(AcpiPciHpState *s, PCIBus *root_bus,
- MemoryRegion *address_space_io)
+ MemoryRegion *address_space_io, bool bridges_enabled)
{
s-root= root_bus;
+s-legacy_piix = !bridges_enabled;
memory_region_init_io(s-io, NULL, acpi_pcihp_io_ops, s,
acpi-pci-hotplug,
PCI_HOTPLUG_SIZE);
diff --git a/hw/acpi/piix4.c b/hw/acpi/piix4.c
index 5d55a3c..2aedfe5 100644
--- a/hw/acpi/piix4.c
+++ b/hw/acpi/piix4.c
@@ -695,7 +695,8 @@ static void piix4_acpi_system_hot_add_init(MemoryRegion
*parent,
memory_region_add_subregion(parent, GPE_BASE, s-io_gpe);
if (s-use_acpi_pci_hotplug) {
-acpi_pcihp_init(s-acpi_pci_hotplug, bus, parent);
+acpi_pcihp_init(s-acpi_pci_hotplug, bus, parent,
+s-use_acpi_pci_hotplug);
} else {
memory_region_init_io(s-io_pci, OBJECT(s), piix4_pci_ops, s,
acpi-pci-hotplug, PCI_HOTPLUG_SIZE);
diff --git a/include/hw/acpi/pcihp.h b/include/hw/acpi/pcihp.h
index aa297c2..02d3ce3 100644
--- a/include/hw/acpi/pcihp.h
+++ b/include/hw/acpi/pcihp.h
@@ -46,10 +46,11 @@ typedef struct AcpiPciHpState {
uint32_t hotplug_select;
PCIBus *root;
MemoryRegion io;
+bool legacy_piix;
} AcpiPciHpState;
void acpi_pcihp_init(AcpiPciHpState *, PCIBus *root,
- MemoryRegion *address_space_io);
+ MemoryRegion *address_space_io, bool bridges_enabled);
/* Invoke on device hotplug */
int acpi_pcihp_device_hotplug(AcpiPciHpState *, PCIDevice *,
--
1.7.1
[Qemu-devel] [PATCH v4 0/5] pc: make ACPI pcihp more reusable
changes since v3: - keep acpi-pci-hotplug-with-bridge-support property and pass its value to acpi_pcihp_init() to init internal .legacy_piix field. changes since v2: - rename 'use_acpi_pci_hotplug' field to 'legacy_piix' and corresponding properties - drop excessive checks for legacy mode - rework legacy vmstate handling to use AcpiPciHpPciStatus[0] structure which reduced acpi/piix4.c by another 50 LOC - move legacy initialization to pcihp.c changes since v1: - add BSEL property to root bus when in compatibility mode as result dropped hw:acpi:pcihp: assume root PCI bus if bus has no ACPI_PCIHP_PROP_BSEL property - rebasing on top of pcihp: reduce number of device check events added patch 3/4. Reuses new pcihp code for handling legacy PCI hotplug in acpi/piix4_pm, which significantly reduces code duplication between piix4_pm and pcihp reducing codebase by ~200 LOC. It also allows cleaner rebase of generic hotplug refactoring, by not increasing code duplication since it will need to be done only for pcihp. Git tree for testing based on mst's PCI tree: https://github.com/imammedo/qemu/commits/pcihp_cleanup_v4 Igor Mammedov (5): pcihp: replace enable|disable_device() with oneliners pcihp: make PCI hotplug mmio handlers indifferent to PCI_HOTPLUG_ADDR pcihp: make pci_read() mmio calback compatible with legacy ACPI hotplug pcihp: remove unused AcpiPciHpPciStatus.device_present field hw:piix4:acpi: reuse pcihp code for legacy PCI hotplug hw/acpi/pcihp.c | 68 --- hw/acpi/piix4.c | 210 --- include/hw/acpi/pcihp.h |5 +- 3 files changed, 55 insertions(+), 228 deletions(-)
[Qemu-devel] [PATCH v4 5/5] hw:piix4:acpi: reuse pcihp code for legacy PCI hotplug
reduces acpi PCI hotplug code duplication by ~200LOC
Signed-off-by: Igor Mammedov imamm...@redhat.com
---
v3:
- remove exessive checks for compat mode
- replace up/down fields of pci_status with corresponding fields
form AcpiPciHpPciStatus in vmstate wich allow to remove extra
50 LOC of legacy code
- move legacy initialization logic to pcihp.c
v2:
- replace obsolete 'device_present' with 'up' field
- add/set ACPI_PCIHP_PROP_BSEL to 0 when running in compatibility
mode with old machine types.
---
hw/acpi/pcihp.c | 23 -
hw/acpi/piix4.c | 211 ---
include/hw/acpi/pcihp.h |1 +
3 files changed, 34 insertions(+), 201 deletions(-)
diff --git a/hw/acpi/pcihp.c b/hw/acpi/pcihp.c
index 974f01c..1ce6fc2 100644
--- a/hw/acpi/pcihp.c
+++ b/hw/acpi/pcihp.c
@@ -46,8 +46,9 @@
# define ACPI_PCIHP_DPRINTF(format, ...) do { } while (0)
#endif
-#define PCI_HOTPLUG_ADDR 0xae00
-#define PCI_HOTPLUG_SIZE 0x0014
+#define ACPI_PCIHP_ADDR 0xae00
+#define ACPI_PCIHP_SIZE 0x0014
+#define ACPI_PCIHP_LEGACY_SIZE 0x000f
#define PCI_UP_BASE 0x
#define PCI_DOWN_BASE 0x0004
#define PCI_EJ_BASE 0x0008
@@ -277,12 +278,24 @@ static const MemoryRegionOps acpi_pcihp_io_ops = {
void acpi_pcihp_init(AcpiPciHpState *s, PCIBus *root_bus,
MemoryRegion *address_space_io, bool bridges_enabled)
{
+uint16_t io_size = ACPI_PCIHP_SIZE;
+
s-root= root_bus;
s-legacy_piix = !bridges_enabled;
+
+if (s-legacy_piix) {
+unsigned *bus_bsel = g_malloc(sizeof *bus_bsel);
+
+io_size = ACPI_PCIHP_LEGACY_SIZE;
+
+*bus_bsel = ACPI_PCIHP_BSEL_DEFAULT;
+object_property_add_uint32_ptr(OBJECT(root_bus), ACPI_PCIHP_PROP_BSEL,
+ bus_bsel, NULL);
+}
+
memory_region_init_io(s-io, NULL, acpi_pcihp_io_ops, s,
- acpi-pci-hotplug,
- PCI_HOTPLUG_SIZE);
-memory_region_add_subregion(address_space_io, PCI_HOTPLUG_ADDR, s-io);
+ acpi-pci-hotplug, io_size);
+memory_region_add_subregion(address_space_io, ACPI_PCIHP_ADDR, s-io);
}
const VMStateDescription vmstate_acpi_pcihp_pci_status = {
diff --git a/hw/acpi/piix4.c b/hw/acpi/piix4.c
index 2aedfe5..7a0efcb 100644
--- a/hw/acpi/piix4.c
+++ b/hw/acpi/piix4.c
@@ -44,13 +44,6 @@
#define GPE_BASE 0xafe0
#define GPE_LEN 4
-#define PCI_HOTPLUG_ADDR 0xae00
-#define PCI_HOTPLUG_SIZE 0x000f
-#define PCI_UP_BASE 0xae00
-#define PCI_DOWN_BASE 0xae04
-#define PCI_EJ_BASE 0xae08
-#define PCI_RMV_BASE 0xae0c
-
#define PIIX4_PCI_HOTPLUG_STATUS 2
struct pci_status {
@@ -80,13 +73,6 @@ typedef struct PIIX4PMState {
Notifier machine_ready;
Notifier powerdown_notifier;
-/* for legacy pci hotplug (compatible with qemu 1.6 and older) */
-MemoryRegion io_pci;
-struct pci_status pci0_status;
-uint32_t pci0_hotplug_enable;
-uint32_t pci0_slot_device_present;
-
-/* for new pci hotplug (with PCI2PCI bridge support) */
AcpiPciHpState acpi_pci_hotplug;
bool use_acpi_pci_hotplug;
@@ -170,17 +156,6 @@ static void pm_write_config(PCIDevice *d,
}
}
-static void vmstate_pci_status_pre_save(void *opaque)
-{
-struct pci_status *pci0_status = opaque;
-PIIX4PMState *s = container_of(pci0_status, PIIX4PMState, pci0_status);
-
-/* We no longer track up, so build a safe value for migrating
- * to a version that still does... of course these might get lost
- * by an old buggy implementation, but we try. */
-pci0_status-up = s-pci0_slot_device_present s-pci0_hotplug_enable;
-}
-
static int vmstate_acpi_post_load(void *opaque, int version_id)
{
PIIX4PMState *s = opaque;
@@ -216,10 +191,9 @@ static const VMStateDescription vmstate_pci_status = {
.version_id = 1,
.minimum_version_id = 1,
.minimum_version_id_old = 1,
-.pre_save = vmstate_pci_status_pre_save,
.fields = (VMStateField []) {
-VMSTATE_UINT32(up, struct pci_status),
-VMSTATE_UINT32(down, struct pci_status),
+VMSTATE_UINT32(up, struct AcpiPciHpPciStatus),
+VMSTATE_UINT32(down, struct AcpiPciHpPciStatus),
VMSTATE_END_OF_LIST()
}
};
@@ -256,7 +230,8 @@ static int acpi_load_old(QEMUFile *f, void *opaque, int
version_id)
qemu_get_be16s(f, temp);
}
-ret = vmstate_load_state(f, vmstate_pci_status, s-pci0_status, 1);
+ret = vmstate_load_state(f, vmstate_pci_status,
+s-acpi_pci_hotplug.acpi_pcihp_pci_status[ACPI_PCIHP_BSEL_DEFAULT],
1);
return ret;
}
@@ -294,70 +269,18 @@ static const VMStateDescription vmstate_acpi = {
VMSTATE_TIMER(ar.tmr.timer, PIIX4PMState),
VMSTATE_INT64(ar.tmr.overflow_time, PIIX4PMState),
VMSTATE_STRUCT(ar.gpe, PIIX4PMState, 2, vmstate_gpe, ACPIGPE),
-VMSTATE_STRUCT_TEST(pci0_status, PIIX4PMState,
-
[Qemu-devel] [PATCH v4 4/5] pcihp: remove unused AcpiPciHpPciStatus.device_present field
Remove now unused 'device_present' field wich was obsoleted by
patch pcihp: reduce number of device check events
Signed-off-by: Igor Mammedov imamm...@redhat.com
---
include/hw/acpi/pcihp.h |1 -
1 files changed, 0 insertions(+), 1 deletions(-)
diff --git a/include/hw/acpi/pcihp.h b/include/hw/acpi/pcihp.h
index 02d3ce3..1fd90e1 100644
--- a/include/hw/acpi/pcihp.h
+++ b/include/hw/acpi/pcihp.h
@@ -35,7 +35,6 @@ typedef struct AcpiPciHpPciStatus {
uint32_t up;
uint32_t down;
uint32_t hotplug_enable;
-uint32_t device_present;
} AcpiPciHpPciStatus;
#define ACPI_PCIHP_PROP_BSEL acpi-pcihp-bsel
--
1.7.1
Re: [Qemu-devel] 9pfs troubles (was Re: [PATCH 1/4] hw/9pfs: fix error handing in local_ioc_getversion())
On Mon, Feb 03, 2014 at 03:05:10PM +0530, Aneesh Kumar K.V wrote: Michael S. Tsirkin m...@redhat.com writes: Haven't used 9pfs in a while. I thought these patches are a good time to play with it some more. I have encountered two issues. What I'm doing: host: qemu a75143eda2ddf581b51e96c000974bcdfe2cbd10. /scm/qemu/x86_64-softmmu/qemu-system-x86_64 -enable-kvm -m 1g -cpu kvm64 -smp 2 f20-x64.qcow2 -netdev user,id=foo -redir tcp:8022::22 -device virtio-net,netdev=foo -serial stdio -fsdev local,security_model=none,id=fsdev0,path=/lib/modules/ -device virtio-9p-pci,id=fs0,fsdev=fsdev0,mount_tag=libmodulesshare -fsdev local,security_model=none,id=fsdev1,path=/boot -device virtio-9p-pci,id=fs1,fsdev=fsdev0,mount_tag=bootshare -no-reboot -snapshot guest: Fedora 20 added this in /etc/fstab: bootshare /share/boot 9p trans=virtio,version=9p2000.L 0 0 libmodulesshare /share/lib/modules 9p trans=virtio,version=9p2000.L 0 0 I have encountered two issues: 1. mount failure on boot If I try to mount on boot through fstab, I get: [2.270157] 9pnet: Could not find request transport: virtio [2.270158] 9pnet: Could not find request transport: virtio Missing 9pnet_virtio.ko module ? Maybe it's loaded too late. But when I get to plymouth prompt it's loaded fine. If I then re-try mount, it succeeds immediately! Some kind of dependency issue? 2. files immediately in the mounted directory aren't visible on the guest under /share/boot. For example, files under /boot on host are not visible on guest, files under child directories seem visible. can you share more details on this ? /boot permissions. ls -al output on host etc. -aneesh for /boot: dr-xr-xr-x. 7 root root 12288 Feb 2 23:41 /boot/ $ ls -la total 739740 dr-xr-xr-x. 7 root root12288 Feb 2 23:41 . dr-xr-xr-x. 22 root root 4096 Feb 2 19:16 .. -rw-r--r--. 1 root root 138741 Dec 23 19:19 config-3.12.6-200.fc19.i686 -rw-r--r--. 1 root root 138724 Jan 10 18:06 config-3.12.7-200.fc19.i686 -rw-r--r--. 1 root root 138724 Jan 16 06:43 config-3.12.8-200.fc19.i686 drwxr-xr-x. 3 root root 4096 May 22 2012 efi -rw-r--r--. 1 root root 178176 Sep 16 13:14 elf-memtest86+-4.20 drwxr-xr-x. 2 root root 4096 Sep 8 14:40 extlinux drwxr-xr-x. 2 root root 4096 Jan 24 2013 grub drwxr-xr-x. 6 root root 4096 Feb 2 23:41 grub2 -rw---. 1 root root 25541989 Sep 8 16:12 initramfs-0-rescue-2b6e810f801e4f458fa97f9a3b9c8a3e.img -rw---. 1 root root 7633329 Jul 11 2013 initramfs-3.10.0-mst.img -rw---. 1 root root 7745838 May 30 2013 initramfs-3.10.0-rc3-mst.img -rw---. 1 root root 7629376 Jul 4 2013 initramfs-3.10.0-rc6-mst.img -rw---. 1 root root 7481618 Sep 22 17:21 initramfs-3.11.0-mst.img -rw---. 1 root root 7657392 Aug 7 16:39 initramfs-3.11.0-rc3-mst.img -rw---. 1 root root 7659204 Aug 18 13:21 initramfs-3.11.0-rc5-mst.img -rw---. 1 root root 7658165 Aug 26 11:47 initramfs-3.11.0-rc7-mst.img -rw---. 1 root root 7510290 Nov 20 15:38 initramfs-3.12.0-mst.img -rw---. 1 root root 7277585 Sep 29 15:14 initramfs-3.12.0-rc2-mst.img -rw---. 1 root root 7277729 Oct 31 07:48 initramfs-3.12.0-rc5-mst.img -rw---. 1 root root 8118663 Jan 6 12:31 initramfs-3.12.6-200.fc19.i686.img -rw---. 1 root root 7273738 Jan 18 19:14 initramfs-3.12.7-200.fc19.i686.img -rw---. 1 root root 7272946 Jan 26 16:54 initramfs-3.12.8-200.fc19.i686.img -rw-r--r--. 1 root root 7334316 Feb 2 23:41 initramfs-3.13.0-mst.img -rw-r--r--. 1 root root 13623844 Jul 19 2012 initramfs-3.4.0-test.img -rw-r--r--. 1 root root 14134095 Jun 17 2012 initramfs-3.5.0-rc2.img -rw-r--r--. 1 root root 6195273 Jun 18 2012 initramfs-3.5.0-rc2-mst.img -rw-r--r--. 1 root root 14137528 Jun 25 2012 initramfs-3.5.0-rc4-mst.img -rw-r--r--. 1 root root 14145675 Jul 19 2012 initramfs-3.5.0-rc7-mst.img -rw-r--r--. 1 root root 14219388 Aug 14 2012 initramfs-3.6.0-rc1-mst.img -rw-r--r--. 1 root root 14230447 Sep 5 2012 initramfs-3.6.0-rc3-mst.img -rw-r--r--. 1 root root 14229895 Sep 25 2012 initramfs-3.6.0-rc5-mst.img -rw---. 1 root root 14431655 Nov 1 2012 initramfs-3.7.0-rc1-mst.img -rw---. 1 root root 14432680 Oct 29 2012 initramfs-3.7.0-rc2-mst.img -rw---. 1 root root 14436612 Dec 5 2012 initramfs-3.7.0-rc7-mst.img -rw---. 1 root root 7485502 Feb 28 2013 initramfs-3.8.0-mst.img -rw---. 1 root root 14549268 Jan 7 2013 initramfs-3.8.0-rc2-mst.img -rw---. 1 root root 14544287 Jan 16 2013 initramfs-3.8.0-rc3-mst.img -rw---. 1 root root 7549459 Jan 24 2013 initramfs-3.8.0-rc4-mst.img -rw---. 1 root root 7434071 Feb 7 2013 initramfs-3.8.0-rc5-mst.img -rw---. 1 root root 7433015 Feb 5 2013 initramfs-3.8.0-rc6-mst.img -rw---. 1 root root 7533783 Apr 4 2013 initramfs-3.9.0-rc5-mst.img -rw---. 1 root root 7551614
Re: [Qemu-devel] [PATCH 1/6] cris: Add a CRISv32 default any CPU for user mode emulation
On Sun, Feb 02, 2014 at 03:04:47AM +, edgar.igles...@gmail.com wrote:
From: Edgar E. Iglesias edgar.igles...@xilinx.com
This is not a regression, but my static busybox sample I have fails to
run with -cpu crisv32. -cpu crisv11 works a little better, but most
syscalls will blow up. Now it is entirely possible my cris busybox binary is
broken, as I have no hw to test it...
./cris-linux-user/qemu-cris -cpu crisv32
/home/voipio/linaro/qemu-smoke/cris/busybox echo hi
PC=80138 CCS=e8 btaken=1 btarget=80138
cc_op=8 cc_src=4 cc_dest=-150999004 cc_result=f6fff020 cc_mask=f
$r0=001dca60 $r1= $r2= $r3=
$r4= $r5= $r6= $r7=
$r8= $r9=0012dc50 $r10=000806a6 $r11=0003
$r12=f6fff02c $r13=1966 $sp=f6fff020 $acr=f6fff01c
special regs:
$bz= $vr=0020 $pid= $srs=
$wz= $exs= $eda=001dca60 $mof=
$dz= $ebp= $erp= $srp=
$nrp= $ccs=00e8 $usp= $spc=
support function regs bank 0:
s00= s01= s02= s03=
s04= s05= s06= s07=
s08= s09= s10= s11=
s12= s13= s14= s15=
Signed-off-by: Edgar E. Iglesias edgar.igles...@xilinx.com
---
target-cris/cpu.c | 7 +++
1 file changed, 7 insertions(+)
diff --git a/target-cris/cpu.c b/target-cris/cpu.c
index 44301a4..21f1860 100644
--- a/target-cris/cpu.c
+++ b/target-cris/cpu.c
@@ -239,7 +239,14 @@ static const TypeInfo cris_cpu_model_type_infos[] = {
.name = TYPE(crisv32),
.parent = TYPE_CRIS_CPU,
.class_init = crisv32_cpu_class_init,
+},
+#if defined(CONFIG_USER_ONLY)
+{
+.name = TYPE(any),
+.parent = TYPE_CRIS_CPU,
+.class_init = crisv32_cpu_class_init,
}
+#endif
};
#undef TYPE
--
1.8.3.2
Re: [Qemu-devel] [PATCH 0/2] qtest: don't leak pid files and UNIX domain sockets
Am 03.02.2014 10:54, schrieb Stefan Hajnoczi: On Fri, Jan 31, 2014 at 12:07:34AM +, Peter Maydell wrote: On 21 November 2013 11:03, Stefan Hajnoczi stefa...@redhat.com wrote: GLib uses abort(3) to exit failed test cases. As a result, the pid file and UNIX domain sockets for a running test are leaked upon failure. Since abort(3) does not call atexit(3) handler functions, we could set up a SIGABRT handler that performs cleanup. But there are other conditions where processes die, like SIGSEGV or SIGBUS. Let's unlink pid files and UNIX domain sockets as soon as the QEMU process has initialized and connections have been made. This eliminates the possibility of leaking these files. So looking back through mailing list history suggests that these patches are supposed to avoid intermittent make check failures like: TEST: tests/qom-test... (pid=5078) /i386/qom/none: ** ERROR:/home/petmay01/linaro/qemu-for-merges/tests/libqtest.c:71:init_socket: assertion failed (ret != -1): (-1 != -1) FAIL GTester: last random seed: R02S79ea313790bc9a8b21d9af5ed55c2fff (pid=5080) /i386/qom/pc:OK /i386/qom/isapc: OK /i386/qom/q35: OK FAIL: tests/qom-test but this patch series doesn't actually say that's what it's for, so does it fix that kind of error? I still think we should merge these patches :). +1 As an explanation, the temporary files contain the PID. When they remain behind due to test failure *and* the PID wraps around and file names thus happen to match, the error was triggered, and thereby not on each run but seemingly sometimes. I am not 100% familiar with the unlinking and code ordering here, but it had looked sane to me back when I looked at it, I just didn't feel confident enough for a Reviewed-by. I could give it a spin and add a Tested-by if that reassures PMM. Andreas Are you happy to merge them? Stefan -- SUSE LINUX Products GmbH, Maxfeldstr. 5, 90409 Nürnberg, Germany GF: Jeff Hawn, Jennifer Guild, Felix Imendörffer; HRB 16746 AG Nürnberg
Re: [Qemu-devel] [PATCH 0/2] qtest: don't leak pid files and UNIX domain sockets
On 3 February 2014 11:10, Andreas Färber afaer...@suse.de wrote: As an explanation, the temporary files contain the PID. When they remain behind due to test failure *and* the PID wraps around and file names thus happen to match, the error was triggered, and thereby not on each run but seemingly sometimes. Yeah, that's what it looks like to me too. I am not 100% familiar with the unlinking and code ordering here, but it had looked sane to me back when I looked at it, I just didn't feel confident enough for a Reviewed-by. I could give it a spin and add a Tested-by if that reassures PMM. I'm happy the code is good, I'm just being lazy about directly applying patches to master, especially since Anthony's patches db is currently stalled and not updating :-) (I guess I need to deal with that some time, though, since we don't have a mechanism for handling patches that fall through the cracks between subsystems beyond somebody with commit access applies them...) thanks -- PMM
Re: [Qemu-devel] [PATCH 1/6] cris: Add a CRISv32 default any CPU for user mode emulation
Hi,
Am 02.02.2014 04:04, schrieb edgar.igles...@gmail.com:
From: Edgar E. Iglesias edgar.igles...@xilinx.com
Signed-off-by: Edgar E. Iglesias edgar.igles...@xilinx.com
---
target-cris/cpu.c | 7 +++
1 file changed, 7 insertions(+)
diff --git a/target-cris/cpu.c b/target-cris/cpu.c
index 44301a4..21f1860 100644
--- a/target-cris/cpu.c
+++ b/target-cris/cpu.c
@@ -239,7 +239,14 @@ static const TypeInfo cris_cpu_model_type_infos[] = {
.name = TYPE(crisv32),
.parent = TYPE_CRIS_CPU,
.class_init = crisv32_cpu_class_init,
+},
+#if defined(CONFIG_USER_ONLY)
+{
+.name = TYPE(any),
+.parent = TYPE_CRIS_CPU,
+.class_init = crisv32_cpu_class_init,
}
If this model is supposed to be exactly the same as crisv32, I would
recommend to make this an alias handled in CRISCPU::class_by_name()
rather than a distinct QOM type.
Regards,
Andreas
+#endif
};
#undef TYPE
--
SUSE LINUX Products GmbH, Maxfeldstr. 5, 90409 Nürnberg, Germany
GF: Jeff Hawn, Jennifer Guild, Felix Imendörffer; HRB 16746 AG Nürnberg
Re: [Qemu-devel] [PATCH V10 06/13] quorum: Add quorum mechanism.
Le Saturday 01 Feb 2014 à 00:04:01 (+0100), Max Reitz a écrit :
On 28.01.2014 17:52, Benoît Canet wrote:
From: Benoît Canet ben...@irqsave.net
Use gnutls's SHA-256 to compare versions.
Signed-off-by: Benoit Canet ben...@irqsave.net
---
block/Makefile.objs | 2 +-
block/quorum.c| 333
+-
configure | 36 +
docs/qmp/qmp-events.txt | 33 +
include/monitor/monitor.h | 2 +
monitor.c | 2 +
6 files changed, 406 insertions(+), 2 deletions(-)
diff --git a/block/Makefile.objs b/block/Makefile.objs
index a2650b9..4ca9d43 100644
--- a/block/Makefile.objs
+++ b/block/Makefile.objs
@@ -3,7 +3,7 @@ block-obj-y += qcow2.o qcow2-refcount.o qcow2-cluster.o
qcow2-snapshot.o qcow2-c
block-obj-y += qed.o qed-gencb.o qed-l2-cache.o qed-table.o qed-cluster.o
block-obj-y += qed-check.o
block-obj-$(CONFIG_VHDX) += vhdx.o vhdx-endian.o vhdx-log.o
-block-obj-y += quorum.o
+block-obj-$(CONFIG_QUORUM) += quorum.o
block-obj-y += parallels.o blkdebug.o blkverify.o
block-obj-y += snapshot.o qapi.o
block-obj-$(CONFIG_WIN32) += raw-win32.o win32-aio.o
diff --git a/block/quorum.c b/block/quorum.c
index 5bf37b3..c319719 100644
--- a/block/quorum.c
+++ b/block/quorum.c
@@ -13,7 +13,43 @@
* See the COPYING file in the top-level directory.
*/
+#include gnutls/gnutls.h
+#include gnutls/crypto.h
#include block/block_int.h
+#include qapi/qmp/qjson.h
+
+#define HASH_LENGTH 32
+
+/* This union holds a vote hash value */
+typedef union QuorumVoteValue {
+char h[HASH_LENGTH]; /* SHA-256 hash */
+int64_t l; /* simpler 64 bits hash */
+} QuorumVoteValue;
+
+/* A vote item */
+typedef struct QuorumVoteItem {
+int index;
+QLIST_ENTRY(QuorumVoteItem) next;
+} QuorumVoteItem;
+
+/* this structure is a vote version. A version is the set of votes sharing
the
+ * same vote value.
+ * The set of votes will be tracked with the items field and its
cardinality is
+ * vote_count.
+ */
+typedef struct QuorumVoteVersion {
+QuorumVoteValue value;
+int index;
+int vote_count;
+QLIST_HEAD(, QuorumVoteItem) items;
+QLIST_ENTRY(QuorumVoteVersion) next;
+} QuorumVoteVersion;
+
+/* this structure holds a group of vote versions together */
+typedef struct QuorumVotes {
+QLIST_HEAD(, QuorumVoteVersion) vote_list;
+int (*compare)(QuorumVoteValue *a, QuorumVoteValue *b);
+} QuorumVotes;
/* the following structure holds the state of one quorum instance */
typedef struct {
@@ -60,10 +96,14 @@ struct QuorumAIOCB {
int success_count; /* number of successfully completed AIOCB
*/
bool *finished; /* completion signal for cancel */
+QuorumVotes votes;
+
bool is_read;
int vote_ret;
};
+static void quorum_vote(QuorumAIOCB *acb);
+
static void quorum_aio_cancel(BlockDriverAIOCB *blockacb)
{
QuorumAIOCB *acb = container_of(blockacb, QuorumAIOCB, common);
@@ -111,6 +151,10 @@ static void quorum_aio_finalize(QuorumAIOCB *acb)
acb-aios[i].ret = 0;
}
+if (acb-vote_ret) {
+ret = acb-vote_ret;
+}
Hm, this makes the vote_ret take precedence over other errors
returned by the children. If they differ (and both are not 0), we
can choose between returning either (vote_ret or “real” errors
reported by the child devices). Both are errors, so I don't see any
natural precedence. However, generally, vote_ret will contain a more
generic error code (i.e., -EIO), thus I could imagine the other
error code reported by the child device to be more appropriate, as
it might contain more useful information.
But, well, on the other had, Quorum is designed for hiding errors
reported by a minority of child devices; therefore, hiding such
errors in this case as well is probably just consequent.
Yes the general idea of quorum is to hide errors as long as the majority is
reached.
For the case where too many children errors occured to be able to do a vote
there is:
ret = s-threshold = acb-success_count ? 0 : quorum_get_first_error(acb);
Kevin asked me do return the first error instead of -EIO.
I'll leave this up to you; I'm fine with it as it is and I'd be fine
if (for whatever reason) you were to change it. :-)
+
acb-common.cb(acb-common.opaque, ret);
if (acb-finished) {
*acb-finished = true;
@@ -122,6 +166,11 @@ static void quorum_aio_finalize(QuorumAIOCB *acb)
qemu_aio_release(acb);
}
+static int quorum_sha256_compare(QuorumVoteValue *a, QuorumVoteValue *b)
+{
+return memcmp(a-h, b-h, HASH_LENGTH);
+}
+
static QuorumAIOCB *quorum_aio_get(BDRVQuorumState *s,
BlockDriverState *bs,
QEMUIOVector *qiov,
@@ -141,6 +190,7 @@ static QuorumAIOCB
Re: [Qemu-devel] [PATCH V10 09/13] quorum: Add quorum_co_get_block_status.
Le Sunday 02 Feb 2014 à 22:44:07 (+0100), Max Reitz a écrit :
On 28.01.2014 17:52, Benoît Canet wrote:
From: Benoît Canet ben...@irqsave.net
Signed-off-by: Benoit Canet ben...@irqsave.net
---
block/quorum.c | 67
++
1 file changed, 67 insertions(+)
diff --git a/block/quorum.c b/block/quorum.c
index a47cd33..9b0718b 100644
--- a/block/quorum.c
+++ b/block/quorum.c
@@ -171,6 +171,22 @@ static int quorum_sha256_compare(QuorumVoteValue *a,
QuorumVoteValue *b)
return memcmp(a-h, b-h, HASH_LENGTH);
}
+static int quorum_64bits_compare(QuorumVoteValue *a, QuorumVoteValue *b)
+{
+int64_t i = a-l;
+int64_t j = b-l;
+
+if (i j) {
+return -1;
+}
+
+if (i j) {
+return 1;
+}
+
+return 0;
+}
+
static QuorumAIOCB *quorum_aio_get(BDRVQuorumState *s,
BlockDriverState *bs,
QEMUIOVector *qiov,
@@ -587,6 +603,56 @@ static void quorum_invalidate_cache(BlockDriverState
*bs)
}
}
+static int64_t coroutine_fn quorum_co_get_block_status(BlockDriverState *bs,
+ int64_t sector_num,
+ int nb_sectors,
+ int *pnum)
+{
+BDRVQuorumState *s = bs-opaque;
+QuorumVoteVersion *winner = NULL;
+QuorumVotes result_votes, num_votes;
+QuorumVoteValue result_value, num_value;
+int i, num;
+int64_t result = 0;
+
+QLIST_INIT(result_votes.vote_list);
+QLIST_INIT(num_votes.vote_list);
+result_votes.compare = quorum_64bits_compare;
+num_votes.compare = quorum_64bits_compare;
+
+for (i = 0; i s-total; i++) {
+result = bdrv_get_block_status(s-bs[i], sector_num, nb_sectors,
num);
+/* skip failed requests */
+if (result 0) {
+continue;
+}
+result_value.l = result BDRV_BLOCK_DATA;
+num_value.l = num;
+quorum_count_vote(result_votes, result_value, i);
+quorum_count_vote(num_votes, num_value, i);
+}
+
+winner = quorum_get_vote_winner(result_votes);
+result = winner-value.l;
Below, you're reading the winning value after checking whether it's
corresponding votes exceeded the threshold. It doesn't matter in the
end, but for the sake of uniformity, I'd do it the same way here
(i.e., move this statement below the if block).
+if (winner-vote_count s-threshold) {
+result = -ERANGE;
Is there any specific reason why you're returning -ERANGE here and
-EIO everywhere else (even in quorum_getlength())?
I probably though It's a comparison so it return -ERANGE.
A bad reason so :(
Max
+goto free_exit;
+}
+
+winner = quorum_get_vote_winner(num_votes);
+if (winner-vote_count s-threshold) {
+result = -ERANGE;
+goto free_exit;
+}
+*pnum = winner-value.l;
+
+free_exit:
+quorum_free_vote_list(result_votes);
+quorum_free_vote_list(num_votes);
+
+return result;
+}
+
static BlockDriver bdrv_quorum = {
.format_name= quorum,
.protocol_name = quorum,
@@ -598,6 +664,7 @@ static BlockDriver bdrv_quorum = {
.bdrv_aio_readv = quorum_aio_readv,
.bdrv_aio_writev= quorum_aio_writev,
.bdrv_invalidate_cache = quorum_invalidate_cache,
+.bdrv_co_get_block_status = quorum_co_get_block_status,
};
static void bdrv_quorum_init(void)
Re: [Qemu-devel] [PATCH] hw/xtensa: add support for ML605 and KC705 FPGA board
Am 02.02.2014 03:31, schrieb Max Filippov: Signed-off-by: Max Filippov jcmvb...@gmail.com --- hw/xtensa/xtensa_lx60.c | 51 + 1 file changed, 47 insertions(+), 4 deletions(-) While I can't check the new numbers, the refactoring looks sane, Reviewed-by: Andreas Färber afaer...@suse.de Andreas -- SUSE LINUX Products GmbH, Maxfeldstr. 5, 90409 Nürnberg, Germany GF: Jeff Hawn, Jennifer Guild, Felix Imendörffer; HRB 16746 AG Nürnberg
Re: [Qemu-devel] [PATCH V10 10/13] quorum: Add quorum_co_flush().
Le Sunday 02 Feb 2014 à 23:02:57 (+0100), Max Reitz a écrit :
On 28.01.2014 17:52, Benoît Canet wrote:
From: Benoît Canet ben...@irqsave.net
Makes a vote to select error if any.
Signed-off-by: Benoit Canet ben...@irqsave.net
---
block/quorum.c | 34 ++
1 file changed, 34 insertions(+)
diff --git a/block/quorum.c b/block/quorum.c
index 9b0718b..1b84b07 100644
--- a/block/quorum.c
+++ b/block/quorum.c
@@ -653,12 +653,46 @@ free_exit:
return result;
}
+static coroutine_fn int quorum_co_flush(BlockDriverState *bs)
+{
+BDRVQuorumState *s = bs-opaque;
+QuorumVoteVersion *winner = NULL;
+QuorumVotes error_votes;
+QuorumVoteValue result_value;
+int i;
+int result = 0;
+bool error = false;
+
+QLIST_INIT(error_votes.vote_list);
+error_votes.compare = quorum_64bits_compare;
+
+for (i = 0; i s-total; i++) {
+result = bdrv_co_flush(s-bs[i]);
+if (result) {
+error = true;
+result_value.l = result;
+quorum_count_vote(error_votes, result_value, i);
+}
+}
+
+if (error) {
+winner = quorum_get_vote_winner(error_votes);
+result = winner-value.l;
+}
+
+quorum_free_vote_list(error_votes);
+
+return result;
+}
+
static BlockDriver bdrv_quorum = {
.format_name= quorum,
.protocol_name = quorum,
.instance_size = sizeof(BDRVQuorumState),
+.bdrv_co_flush_to_disk = quorum_co_flush,
+
.bdrv_getlength = quorum_getlength,
.bdrv_aio_readv = quorum_aio_readv,
So, my general opinion on this patch (for reads/writes we don't vote
on the error code either; so why here?) hasn't changed, but well, I
definitely don't oppose it.
For the reads/writes Kevin asked me to return the first error.
For the flush function someone else (probably Eric) asked me to do voting.
Another problem, however: If any error occurs, this function will
return an error as well. Is that intended? If an error on a
read/write operation occurs but there are still enough successful
reads/writes to reach quorum, no error is returned. Is there a
reason why this should be different for flush?
If an error occurs multiple times and is able to establish quorum this winning
error will be returned.
Else 0 will be returned due to the vote result if the error is in minority.
So the behavior is similar as long quorum is reached.
Max
Re: [Qemu-devel] [BUGFIX][PATCH v2] configure: Disable libtool if -fPIE does not work with it (bug #1257099)
On Wed, 15 Jan 2014, Paolo Bonzini wrote:
Il 03/01/2014 03:12, Don Slutz ha scritto:
Adjust TMPO and added TMPB, TMPL, and TMPA. libtool needs the names
to be fixed (TMPB).
Add new functions do_libtool and libtool_prog.
Add check for broken gcc and libtool.
Signed-off-by: Don Slutz dsl...@verizon.com
---
Was posted as an attachment.
https://lists.gnu.org/archive/html/qemu-devel/2013-12/msg02678.html
configure | 63
++-
1 file changed, 62 insertions(+), 1 deletion(-)
diff --git a/configure b/configure
index edfea95..852d021 100755
--- a/configure
+++ b/configure
@@ -12,7 +12,10 @@ else
fi
TMPC=${TMPDIR1}/qemu-conf-${RANDOM}-$$-${RANDOM}.c
-TMPO=${TMPDIR1}/qemu-conf-${RANDOM}-$$-${RANDOM}.o
+TMPB=qemu-conf-${RANDOM}-$$-${RANDOM}
+TMPO=${TMPDIR1}/${TMPB}.o
+TMPL=${TMPDIR1}/${TMPB}.lo
+TMPA=${TMPDIR1}/lib${TMPB}.la
TMPE=${TMPDIR1}/qemu-conf-${RANDOM}-$$-${RANDOM}.exe
# NB: do not call exit in the trap handler; this is buggy with some
shells;
@@ -86,6 +89,38 @@ compile_prog() {
do_cc $QEMU_CFLAGS $local_cflags -o $TMPE $TMPC $LDFLAGS $local_ldflags
}
+do_libtool() {
+local mode=$1
+shift
+# Run the compiler, capturing its output to the log.
+echo $libtool $mode --tag=CC $cc $@ config.log
+$libtool $mode --tag=CC $cc $@ config.log 21 || return $?
+# Test passed. If this is an --enable-werror build, rerun
+# the test with -Werror and bail out if it fails. This
+# makes warning-generating-errors in configure test code
+# obvious to developers.
+if test $werror != yes; then
+return 0
+fi
+# Don't bother rerunning the compile if we were already using -Werror
+case $* in
+*-Werror*)
+ return 0
+;;
+esac
+echo $libtool $mode --tag=CC $cc -Werror $@ config.log
+$libtool $mode --tag=CC $cc -Werror $@ config.log 21 return
$?
+error_exit configure test passed without -Werror but failed with
-Werror. \
+This is probably a bug in the configure script. The failing
command \
+will be at the bottom of config.log. \
+You can run configure with --disable-werror to bypass this check.
+}
+
+libtool_prog() {
+do_libtool --mode=compile $QEMU_CFLAGS -c -fPIE -DPIE -o $TMPO $TMPC
|| return $?
+do_libtool --mode=link $LDFLAGS -o $TMPA $TMPL -rpath /usr/local/lib
+}
+
# symbolically link $1 to $2. Portable version of ln -sf.
symlink() {
rm -rf $2
@@ -1367,6 +1402,32 @@ EOF
fi
fi
+# check for broken gcc and libtool in RHEL5
+if test -n $libtool -a $pie != no ; then
+ cat $TMPC EOF
+
+void *f(unsigned char *buf, int len);
+void *g(unsigned char *buf, int len);
+
+void *
+f(unsigned char *buf, int len)
+{
+return (void*)0L;
+}
+
+void *
+g(unsigned char *buf, int len)
+{
+return f(buf, len);
+}
+
+EOF
+ if ! libtool_prog; then
+echo Disabling libtool due to broken toolchain support
+libtool=
+ fi
+fi
+
##
# __sync_fetch_and_and requires at least -march=i486. Many toolchains
# use i686 as default anyway, but for those that don't, an explicit
I'm applying this to a configure branch on my github repository. Thanks!
Paolo, did this patch ever make it upstream? If so, do you have a commit
id?
Re: [Qemu-devel] [PATCH 1/6] cris: Add a CRISv32 default any CPU for user mode emulation
On Mon, Feb 03, 2014 at 01:02:34PM +0200, Riku Voipio wrote:
On Sun, Feb 02, 2014 at 03:04:47AM +, edgar.igles...@gmail.com wrote:
From: Edgar E. Iglesias edgar.igles...@xilinx.com
This is not a regression, but my static busybox sample I have fails to
run with -cpu crisv32. -cpu crisv11 works a little better, but most
syscalls will blow up. Now it is entirely possible my cris busybox binary is
broken, as I have no hw to test it...
Hi Riku,
The CRISv10 class of CPUs (8, 9, 10, 11) are very different from the
CRISv32. There is some binary compatibility and more ASM compat but
they should really be seen as different archs.
Not sure what kind of binary you have but if you want I can provide
busybox for both v10 and v32.
Cheers,
Edgar
./cris-linux-user/qemu-cris -cpu crisv32
/home/voipio/linaro/qemu-smoke/cris/busybox echo hi
PC=80138 CCS=e8 btaken=1 btarget=80138
cc_op=8 cc_src=4 cc_dest=-150999004 cc_result=f6fff020 cc_mask=f
$r0=001dca60 $r1= $r2= $r3=
$r4= $r5= $r6= $r7=
$r8= $r9=0012dc50 $r10=000806a6 $r11=0003
$r12=f6fff02c $r13=1966 $sp=f6fff020 $acr=f6fff01c
special regs:
$bz= $vr=0020 $pid= $srs=
$wz= $exs= $eda=001dca60 $mof=
$dz= $ebp= $erp= $srp=
$nrp= $ccs=00e8 $usp= $spc=
support function regs bank 0:
s00= s01= s02= s03=
s04= s05= s06= s07=
s08= s09= s10= s11=
s12= s13= s14= s15=
Signed-off-by: Edgar E. Iglesias edgar.igles...@xilinx.com
---
target-cris/cpu.c | 7 +++
1 file changed, 7 insertions(+)
diff --git a/target-cris/cpu.c b/target-cris/cpu.c
index 44301a4..21f1860 100644
--- a/target-cris/cpu.c
+++ b/target-cris/cpu.c
@@ -239,7 +239,14 @@ static const TypeInfo cris_cpu_model_type_infos[] = {
.name = TYPE(crisv32),
.parent = TYPE_CRIS_CPU,
.class_init = crisv32_cpu_class_init,
+},
+#if defined(CONFIG_USER_ONLY)
+{
+.name = TYPE(any),
+.parent = TYPE_CRIS_CPU,
+.class_init = crisv32_cpu_class_init,
}
+#endif
};
#undef TYPE
--
1.8.3.2
Re: [Qemu-devel] [PATCH V10 12/13] quorum: Add quorum_open() and quorum_close().
Le Monday 03 Feb 2014 à 00:09:28 (+0100), Max Reitz a écrit :
On 28.01.2014 17:52, Benoît Canet wrote:
From: Benoît Canet ben...@irqsave.net
Example of command line:
-drive if=virtio,file.driver=quorum,\
file.children.0.file.filename=1.raw,\
file.children.0.node-name=1.raw,\
file.children.0.driver=raw,\
file.children.1.file.filename=2.raw,\
file.children.1.node-name=2.raw,\
file.children.1.driver=raw,\
file.children.2.file.filename=3.raw,\
file.children.2.node-name=3.raw,\
file.children.2.driver=raw,\
file.vote_threshold=2
file.blkverify=on with file.vote_threshold=2 and two files can be passed to
emulated blkverify.
Signed-off-by: Benoit Canet ben...@irqsave.net
---
block/quorum.c | 308
+++
qapi-schema.json | 21 +++-
2 files changed, 328 insertions(+), 1 deletion(-)
diff --git a/block/quorum.c b/block/quorum.c
index e7b2090..0c0d630 100644
--- a/block/quorum.c
+++ b/block/quorum.c
@@ -17,8 +17,12 @@
#include gnutls/crypto.h
#include block/block_int.h
#include qapi/qmp/qjson.h
+#include qapi/qmp/types.h
+#include qemu-common.h
#define HASH_LENGTH 32
+#define KEY_PREFIX children.
+#define KEY_FILENAME_SUFFIX .file.filename
/* This union holds a vote hash value */
typedef union QuorumVoteValue {
@@ -702,12 +706,316 @@ static bool
quorum_recurse_is_first_non_filter(BlockDriverState *bs,
return false;
}
Perhaps you could make use of qdict_extract_subqdict() and
qdict_array_split() functions here...? That is,
qdict_extract_subqdict(..., children.) and then
qdict_array_split() on the result?
+static int quorum_match_key(const char *key,
+char **key_prefix)
+{
+const char *start;
+char *next;
+unsigned long long idx;
+int ret;
+
+*key_prefix = NULL;
+
+/* the following code is a translation of the following pseudo code:
+ * match = key.match('(^children\.(\d+)\.)$suffix)
+ * if not match:
+ * return -1;
+ * key_prefix = match.outer_match()
+ * idx = match.inner_match()
+ *
+ * note: we also match the .file suffix to avoid futher checkings
+ */
+
+/* this key is not a child */
+if (strncmp(key, KEY_PREFIX, strlen(KEY_PREFIX))) {
+return -1;
+}
+
+/* take first char after prefix */
+start = key + strlen(KEY_PREFIX);
+
+/* if the string end here - scan fail */
+if (start[0] == '\0') {
+return -1;
+}
+
+/* try to retrieve the index */
+ret = parse_uint(start, idx, next, 10);
+
+/* no int found - scan fail */
+if (ret 0) {
+return -1;
+}
+
+/* we are taking a reference via QMP */
+if (next - key == strlen(key)) {
if (!*next) would probably accomplish the same (or if next[0] ==
'\0') to keep your coding style) without having to call strlen().
+*key_prefix = g_strdup(key);
+return idx;
+}
+
+/* match the suffix to avoid matching the same idx
+ * multiple times and be required to do more checks later
+ */
+if (strncmp(next, KEY_FILENAME_SUFFIX, strlen(KEY_FILENAME_SUFFIX))) {
As stated in my review from October, you may use sizeof() to avoid strlen().
+return -1;
+}
+
+/* do not include '.' */
+int len = next - key;
+*key_prefix = g_strndup(key, len);
+
+return idx;
+}
+
+static QDict *quorum_get_children_idx(const QDict *options)
+{
+const QDictEntry *ent;
+QDict *result;
+char *key_prefix;
+int idx;
+
+result = qdict_new();
+
+for (ent = qdict_first(options); ent; ent = qdict_next(options, ent)) {
+const char *key = qdict_entry_key(ent);
+idx = quorum_match_key(key,
+ key_prefix);
+
+/* if the result zero or positive we got a key */
+if (idx 0) {
+continue;
+}
+
+qdict_put(result, key_prefix, qint_from_int(idx));
+}
+
+return result;
+}
+
+static int quorum_fill_validation_array(bool *array,
+const QDict *dict,
+int total,
+Error **errp)
+{
+const QDictEntry *ent;
+
+/* fill the checking array with children indexes */
+for (ent = qdict_first(dict); ent; ent = qdict_next(dict, ent)) {
+const char *key = qdict_entry_key(ent);
+int idx = qdict_get_int(dict, key);
+
+if (idx 0 || idx = total) {
+error_setg(errp,
+Children index must be between 0 and children count -1);
Quote from October: I'd prefer - 1 instead of -1 (or even Child
index must be an unsigned integer smaller than the children count).
+return -ERANGE;
+}
+
+array[idx] = true;
+}
+
+return 0;
+}
+
Re: [Qemu-devel] [PATCH 1/2] qom-test: Run for all available machines
Andreas Färber afaer...@suse.de writes: Am 10.01.2014 14:31, schrieb arm...@redhat.com: From: Markus Armbruster arm...@redhat.com Get available machines via QMP instead of hardcoding a list that's perpetually out of date. A few machines don't work out of the box: * Several ppcemb machines can't initialize their CPU. * Xen machines can work only when running under the Xen hypervisor. Blacklist them. Signed-off-by: Markus Armbruster arm...@redhat.com I've rebased this, queing it without the ppcemb blacklist (and adjusted commit message) while waiting for Alex' ppc pull. Your rebase is fine. Thanks!
Re: [Qemu-devel] [PATCH V10 12/13] quorum: Add quorum_open() and quorum_close().
Le Monday 03 Feb 2014 à 00:09:28 (+0100), Max Reitz a écrit :
On 28.01.2014 17:52, Benoît Canet wrote:
From: Benoît Canet ben...@irqsave.net
Example of command line:
-drive if=virtio,file.driver=quorum,\
file.children.0.file.filename=1.raw,\
file.children.0.node-name=1.raw,\
file.children.0.driver=raw,\
file.children.1.file.filename=2.raw,\
file.children.1.node-name=2.raw,\
file.children.1.driver=raw,\
file.children.2.file.filename=3.raw,\
file.children.2.node-name=3.raw,\
file.children.2.driver=raw,\
file.vote_threshold=2
file.blkverify=on with file.vote_threshold=2 and two files can be passed to
emulated blkverify.
Signed-off-by: Benoit Canet ben...@irqsave.net
---
block/quorum.c | 308
+++
qapi-schema.json | 21 +++-
2 files changed, 328 insertions(+), 1 deletion(-)
diff --git a/block/quorum.c b/block/quorum.c
index e7b2090..0c0d630 100644
--- a/block/quorum.c
+++ b/block/quorum.c
@@ -17,8 +17,12 @@
#include gnutls/crypto.h
#include block/block_int.h
#include qapi/qmp/qjson.h
+#include qapi/qmp/types.h
+#include qemu-common.h
#define HASH_LENGTH 32
+#define KEY_PREFIX children.
+#define KEY_FILENAME_SUFFIX .file.filename
/* This union holds a vote hash value */
typedef union QuorumVoteValue {
@@ -702,12 +706,316 @@ static bool
quorum_recurse_is_first_non_filter(BlockDriverState *bs,
return false;
}
As a general stance sorry to have missed this mail of review from october.
It was not done on purpose.
Best regards
Benoît
Perhaps you could make use of qdict_extract_subqdict() and
qdict_array_split() functions here...? That is,
qdict_extract_subqdict(..., children.) and then
qdict_array_split() on the result?
+static int quorum_match_key(const char *key,
+char **key_prefix)
+{
+const char *start;
+char *next;
+unsigned long long idx;
+int ret;
+
+*key_prefix = NULL;
+
+/* the following code is a translation of the following pseudo code:
+ * match = key.match('(^children\.(\d+)\.)$suffix)
+ * if not match:
+ * return -1;
+ * key_prefix = match.outer_match()
+ * idx = match.inner_match()
+ *
+ * note: we also match the .file suffix to avoid futher checkings
+ */
+
+/* this key is not a child */
+if (strncmp(key, KEY_PREFIX, strlen(KEY_PREFIX))) {
+return -1;
+}
+
+/* take first char after prefix */
+start = key + strlen(KEY_PREFIX);
+
+/* if the string end here - scan fail */
+if (start[0] == '\0') {
+return -1;
+}
+
+/* try to retrieve the index */
+ret = parse_uint(start, idx, next, 10);
+
+/* no int found - scan fail */
+if (ret 0) {
+return -1;
+}
+
+/* we are taking a reference via QMP */
+if (next - key == strlen(key)) {
if (!*next) would probably accomplish the same (or if next[0] ==
'\0') to keep your coding style) without having to call strlen().
+*key_prefix = g_strdup(key);
+return idx;
+}
+
+/* match the suffix to avoid matching the same idx
+ * multiple times and be required to do more checks later
+ */
+if (strncmp(next, KEY_FILENAME_SUFFIX, strlen(KEY_FILENAME_SUFFIX))) {
As stated in my review from October, you may use sizeof() to avoid strlen().
+return -1;
+}
+
+/* do not include '.' */
+int len = next - key;
+*key_prefix = g_strndup(key, len);
+
+return idx;
+}
+
+static QDict *quorum_get_children_idx(const QDict *options)
+{
+const QDictEntry *ent;
+QDict *result;
+char *key_prefix;
+int idx;
+
+result = qdict_new();
+
+for (ent = qdict_first(options); ent; ent = qdict_next(options, ent)) {
+const char *key = qdict_entry_key(ent);
+idx = quorum_match_key(key,
+ key_prefix);
+
+/* if the result zero or positive we got a key */
+if (idx 0) {
+continue;
+}
+
+qdict_put(result, key_prefix, qint_from_int(idx));
+}
+
+return result;
+}
+
+static int quorum_fill_validation_array(bool *array,
+const QDict *dict,
+int total,
+Error **errp)
+{
+const QDictEntry *ent;
+
+/* fill the checking array with children indexes */
+for (ent = qdict_first(dict); ent; ent = qdict_next(dict, ent)) {
+const char *key = qdict_entry_key(ent);
+int idx = qdict_get_int(dict, key);
+
+if (idx 0 || idx = total) {
+error_setg(errp,
+Children index must be between 0 and children count -1);
Quote from October: I'd prefer - 1 instead of -1 (or even Child
index must be an unsigned integer smaller than the
[Qemu-devel] Improving patch tracking - something like gerrit?
Hi all, It should be fairly evident to most people that the volume of patches flowing through the qemu-devel mailing list is continually increasing, and it is becoming increasingly difficult to track which patches have been applied over time. This is particularly a problem where patchsets have dependencies on other patchsets which haven't yet been applied to git master, which can then cause merge conflicts due to length of time taken for the final series to be merged. Is it time for QEMU to start looking at tools such as gerrit to help manage this process? There seems to be an increasing number of ping requests for outstanding patches (including my own) which don't get applied for weeks, and often even months because they target less popular platforms/subsystems and so don't always get the attention of the committers. What I would like to see from such a tool would be something that would enable me to see which patches are being considered for each release, so that I can see if a particular patch I have submitted is being ignored, rejected or deferred to a future release. Note that I think that one of the biggest benefits of such a tool would be during feature freeze, whereby the mailing list contains an avalanche of future and current PULL requests which have to be manually filtered by committers. This would help both developers and committers see what patches are definitely scheduled for the next release as opposed to patches being rejected at the last minute because the PULL request failed just before the release deadline. Does anyone else have any thoughts/ideas as to how to better manage this process, particularly for a project like QEMU where the number of patches is considerably greater than the number of reviewers/committers? ATB, Mark.
Re: [Qemu-devel] [PATCH V10 12/13] quorum: Add quorum_open() and quorum_close().
Le Monday 03 Feb 2014 à 13:21:18 (+0100), Benoît Canet a écrit :
Le Monday 03 Feb 2014 à 00:09:28 (+0100), Max Reitz a écrit :
On 28.01.2014 17:52, Benoît Canet wrote:
From: Benoît Canet ben...@irqsave.net
Example of command line:
-drive if=virtio,file.driver=quorum,\
file.children.0.file.filename=1.raw,\
file.children.0.node-name=1.raw,\
file.children.0.driver=raw,\
file.children.1.file.filename=2.raw,\
file.children.1.node-name=2.raw,\
file.children.1.driver=raw,\
file.children.2.file.filename=3.raw,\
file.children.2.node-name=3.raw,\
file.children.2.driver=raw,\
file.vote_threshold=2
file.blkverify=on with file.vote_threshold=2 and two files can be passed to
emulated blkverify.
Signed-off-by: Benoit Canet ben...@irqsave.net
---
block/quorum.c | 308
+++
qapi-schema.json | 21 +++-
2 files changed, 328 insertions(+), 1 deletion(-)
diff --git a/block/quorum.c b/block/quorum.c
index e7b2090..0c0d630 100644
--- a/block/quorum.c
+++ b/block/quorum.c
@@ -17,8 +17,12 @@
#include gnutls/crypto.h
#include block/block_int.h
#include qapi/qmp/qjson.h
+#include qapi/qmp/types.h
+#include qemu-common.h
#define HASH_LENGTH 32
+#define KEY_PREFIX children.
+#define KEY_FILENAME_SUFFIX .file.filename
/* This union holds a vote hash value */
typedef union QuorumVoteValue {
@@ -702,12 +706,316 @@ static bool
quorum_recurse_is_first_non_filter(BlockDriverState *bs,
return false;
}
Perhaps you could make use of qdict_extract_subqdict() and
qdict_array_split() functions here...? That is,
qdict_extract_subqdict(..., children.) and then
qdict_array_split() on the result?
+static int quorum_match_key(const char *key,
+char **key_prefix)
+{
+const char *start;
+char *next;
+unsigned long long idx;
+int ret;
+
+*key_prefix = NULL;
+
+/* the following code is a translation of the following pseudo code:
+ * match = key.match('(^children\.(\d+)\.)$suffix)
+ * if not match:
+ * return -1;
+ * key_prefix = match.outer_match()
+ * idx = match.inner_match()
+ *
+ * note: we also match the .file suffix to avoid futher checkings
+ */
+
+/* this key is not a child */
+if (strncmp(key, KEY_PREFIX, strlen(KEY_PREFIX))) {
+return -1;
+}
+
+/* take first char after prefix */
+start = key + strlen(KEY_PREFIX);
+
+/* if the string end here - scan fail */
+if (start[0] == '\0') {
+return -1;
+}
+
+/* try to retrieve the index */
+ret = parse_uint(start, idx, next, 10);
+
+/* no int found - scan fail */
+if (ret 0) {
+return -1;
+}
+
+/* we are taking a reference via QMP */
+if (next - key == strlen(key)) {
if (!*next) would probably accomplish the same (or if next[0] ==
'\0') to keep your coding style) without having to call strlen().
+*key_prefix = g_strdup(key);
+return idx;
+}
+
+/* match the suffix to avoid matching the same idx
+ * multiple times and be required to do more checks later
+ */
+if (strncmp(next, KEY_FILENAME_SUFFIX, strlen(KEY_FILENAME_SUFFIX))) {
As stated in my review from October, you may use sizeof() to avoid strlen().
+return -1;
+}
+
+/* do not include '.' */
+int len = next - key;
+*key_prefix = g_strndup(key, len);
+
+return idx;
+}
+
+static QDict *quorum_get_children_idx(const QDict *options)
+{
+const QDictEntry *ent;
+QDict *result;
+char *key_prefix;
+int idx;
+
+result = qdict_new();
+
+for (ent = qdict_first(options); ent; ent = qdict_next(options, ent))
{
+const char *key = qdict_entry_key(ent);
+idx = quorum_match_key(key,
+ key_prefix);
+
+/* if the result zero or positive we got a key */
+if (idx 0) {
+continue;
+}
+
+qdict_put(result, key_prefix, qint_from_int(idx));
+}
+
+return result;
+}
+
+static int quorum_fill_validation_array(bool *array,
+const QDict *dict,
+int total,
+Error **errp)
+{
+const QDictEntry *ent;
+
+/* fill the checking array with children indexes */
+for (ent = qdict_first(dict); ent; ent = qdict_next(dict, ent)) {
+const char *key = qdict_entry_key(ent);
+int idx = qdict_get_int(dict, key);
+
+if (idx 0 || idx = total) {
+error_setg(errp,
+Children index must be between 0 and children count -1);
Quote from
[Qemu-devel] KVM call agenda for 2014-02-04
Hi Please, send any topic that you are interested in covering. * Should we change anything to get more people to sign for the call? There hasn't been a call in quite a long time. Ideas? Thanks, Juan. Call details: 09:00 AM to 10:00 AM EDT Every two weeks If you need phone number details, contact me privately
[Qemu-devel] Query regarding Iotrhead
While debugging using gdb, I found that the interrupt related flow is not handled in io thread i.e. on using info thread, I found: Thread 2(iothread) is at __lll_lock_wait() while Thread 1 is at qemu_set_irq(in irq.c).. Initially I thought that this interrupt related code(kbd_update_irq---qemu_Set_irq..) should be handled through iothread(event loop).. Can u help me regarding this
[Qemu-devel] [PULL 00/01] seccomp: adding new syscalls to the whitelist
The following changes since commit 2f61120c10da9128357510debc8e66880cd2bfdc: Merge remote-tracking branch 'qmp-unstable/queue/qmp' into staging (2014-02-01 23:32:31 +) are available in the git repository at: git://github.com/otubo/qemu.git seccomp Felix Geyer (1): seccomp: add timerfd_create and timerfd_settime to the whitelist qemu-seccomp.c |4 +++- 1 files changed, 3 insertions(+), 1 deletions(-)
[Qemu-devel] [PULL 01/01] seccomp: add kill() to the syscall whitelist
From: Paul Moore pmo...@redhat.com
The kill() syscall is triggered with the following command:
# qemu -sandbox on -monitor stdio \
-device intel-hda -device hda-duplex -vnc :0
The resulting syslog/audit message:
# ausearch -m SECCOMP
time-Wed Nov 20 09:52:08 2013
type=SECCOMP msg=audit(1384912328.482:6656): auid=0 uid=0 gid=0 ses=854
subj=unconfined_u:unconfined_r:unconfined_t:s0-s0:c0.c1023 pid=12087
comm=qemu-kvm sig=31 syscall=62 compat=0 ip=0x7f7a1d2abc67 code=0x0
# scmp_sys_resolver 62
kill
Reported-by: CongLi c...@redhat.com
Tested-by: CongLi c...@redhat.com
Signed-off-by: Paul Moore pmo...@redhat.com
Acked-by: Eduardo Otubo ot...@linux.vnet.ibm.com
---
qemu-seccomp.c |1 +
1 files changed, 1 insertions(+), 0 deletions(-)
diff --git a/qemu-seccomp.c b/qemu-seccomp.c
index 69cee44..cf07869 100644
--- a/qemu-seccomp.c
+++ b/qemu-seccomp.c
@@ -114,6 +114,7 @@ static const struct QemuSeccompSyscall seccomp_whitelist[]
= {
{ SCMP_SYS(write), 244 },
{ SCMP_SYS(fcntl), 243 },
{ SCMP_SYS(tgkill), 242 },
+{ SCMP_SYS(kill), 242 },
{ SCMP_SYS(rt_sigaction), 242 },
{ SCMP_SYS(pipe2), 242 },
{ SCMP_SYS(munmap), 242 },
--
1.7.1
Re: [Qemu-devel] Improving patch tracking - something like gerrit?
On 3 February 2014 12:45, Mark Cave-Ayland mark.cave-ayl...@ilande.co.uk wrote: It should be fairly evident to most people that the volume of patches flowing through the qemu-devel mailing list is continually increasing, and it is becoming increasingly difficult to track which patches have been applied over time. This is particularly a problem where patchsets have dependencies on other patchsets which haven't yet been applied to git master, which can then cause merge conflicts due to length of time taken for the final series to be merged. Is it time for QEMU to start looking at tools such as gerrit to help manage this process? There seems to be an increasing number of ping requests for outstanding patches (including my own) which don't get applied for weeks, and often even months because they target less popular platforms/subsystems and so don't always get the attention of the committers. So in general I would be happy to see better tooling for this (currently I use a combination of gmail folders to track things I ought to be reviewing or pulling plus Anthony's patches tool to do the work of actually applying things). The difficulty is coming up with one of: (a) a system that works even if some maintainers/submaintainers aren't using it (b) a system good enough that we can force everybody to use it (Also (c) somebody needs to put in the work to set the system up on some server, maintain it, explain how it works, etc.) What I would like to see from such a tool would be something that would enable me to see which patches are being considered for each release, so that I can see if a particular patch I have submitted is being ignored, rejected or deferred to a future release. Typically I think submaintainers will reply to patches when they apply them to trees. I suspect your actual problem is there are areas of the code base where patches do just get ignored because of a lack of maintainers for them. You can generally tell the difference between ignored/rejected/deferred with our current system: it's whether there's a reply from somebody with negative review, or saying 'ok but won't go in this release' or no email reply at all. Note that I think that one of the biggest benefits of such a tool would be during feature freeze, whereby the mailing list contains an avalanche of future and current PULL requests which have to be manually filtered by committers. This would help both developers and committers see what patches are definitely scheduled for the next release as opposed to patches being rejected at the last minute because the PULL request failed just before the release deadline. Get pull requests (and patches for the release) in early, please. The closer you get to the release deadline the more likely you are to find that yes, some patches don't make it under the wire. thanks -- PMM
Re: [Qemu-devel] [PATCHv9 1/5] block: add native support for NFS
Am 03.02.2014 um 10:26 hat Peter Lieven geschrieben:
This patch adds native support for accessing images on NFS
shares without the requirement to actually mount the entire
NFS share on the host.
NFS Images can simply be specified by an url of the form:
nfs://host/export/filename[?param=value[param2=value2[...]]]
For example:
qemu-img create -f qcow2 nfs://10.0.0.1/qemu-images/test.qcow2
You need LibNFS from Ronnie Sahlberg available at:
git://github.com/sahlberg/libnfs.git
for this to work.
During configure it is automatically probed for libnfs and support
is enabled on-the-fly. You can forbid or enforce libnfs support
with --disable-libnfs or --enable-libnfs respectively.
Due to NFS restrictions you might need to execute your binaries
as root, allow them to open priviledged ports (1024) or specify
insecure option on the NFS server.
For additional information on ROOT vs. non-ROOT operation and URL
format + parameters see:
https://raw.github.com/sahlberg/libnfs/master/README
Supported by qemu are the uid, gid and tcp-syncnt URL parameters.
LibNFS currently support NFS version 3 only.
Signed-off-by: Peter Lieven p...@kamp.de
Some minor comments in case you need to respin for another reason:
MAINTAINERS |5 +
block/Makefile.objs |1 +
block/nfs.c | 439
+++
configure | 26 +++
qapi-schema.json|1 +
5 files changed, 472 insertions(+)
create mode 100644 block/nfs.c
+typedef struct NFSRPC {
+int ret;
+int complete;
Should be bool.
+QEMUIOVector *iov;
+struct stat *st;
+Coroutine *co;
+QEMUBH *bh;
+} NFSRPC;
+static BlockDriver bdrv_nfs = {
+.format_name = nfs,
+.protocol_name = nfs,
+
+.instance_size = sizeof(NFSClient),
+.bdrv_needs_filename = true,
+.bdrv_has_zero_init = nfs_has_zero_init,
+.bdrv_get_allocated_file_size = nfs_get_allocated_file_size,
+.bdrv_truncate = nfs_file_truncate,
This group of options could have used alignment for = as well. :-)
+.bdrv_file_open = nfs_file_open,
+.bdrv_close = nfs_file_close,
+.bdrv_create = nfs_file_create,
+
+.bdrv_co_readv = nfs_co_readv,
+.bdrv_co_writev= nfs_co_writev,
+.bdrv_co_flush_to_disk = nfs_co_flush,
+};
Kevin
Re: [Qemu-devel] [PATCH 1/6] cris: Add a CRISv32 default any CPU for user mode emulation
On Mon, Feb 03, 2014 at 12:44:03PM +0100, Andreas Färber wrote:
Hi,
Am 02.02.2014 04:04, schrieb edgar.igles...@gmail.com:
From: Edgar E. Iglesias edgar.igles...@xilinx.com
Signed-off-by: Edgar E. Iglesias edgar.igles...@xilinx.com
---
target-cris/cpu.c | 7 +++
1 file changed, 7 insertions(+)
diff --git a/target-cris/cpu.c b/target-cris/cpu.c
index 44301a4..21f1860 100644
--- a/target-cris/cpu.c
+++ b/target-cris/cpu.c
@@ -239,7 +239,14 @@ static const TypeInfo cris_cpu_model_type_infos[] = {
.name = TYPE(crisv32),
.parent = TYPE_CRIS_CPU,
.class_init = crisv32_cpu_class_init,
+},
+#if defined(CONFIG_USER_ONLY)
+{
+.name = TYPE(any),
+.parent = TYPE_CRIS_CPU,
+.class_init = crisv32_cpu_class_init,
}
If this model is supposed to be exactly the same as crisv32, I would
recommend to make this an alias handled in CRISCPU::class_by_name()
rather than a distinct QOM type.
Sounds good, I'll send a v2 of this one.
Cheers,
Edgar
[Qemu-devel] [PATCH v2] cris: Add any as alias for crisv32 in user emulation
From: Edgar E. Iglesias edgar.igles...@xilinx.com
Signed-off-by: Edgar E. Iglesias edgar.igles...@xilinx.com
---
Notes (cl):
v1 - v2:
Implement as an alias rather than as QOM type. (AF)
target-cris/cpu.c | 6 ++
1 file changed, 6 insertions(+)
diff --git a/target-cris/cpu.c b/target-cris/cpu.c
index 44301a4..acb5688 100644
--- a/target-cris/cpu.c
+++ b/target-cris/cpu.c
@@ -66,6 +66,12 @@ static ObjectClass *cris_cpu_class_by_name(const char
*cpu_model)
return NULL;
}
+#if defined(CONFIG_USER_ONLY)
+if (strcasecmp(cpu_model, any) == 0) {
+return object_class_by_name(crisv32- TYPE_CRIS_CPU);
+}
+#endif
+
typename = g_strdup_printf(%s- TYPE_CRIS_CPU, cpu_model);
oc = object_class_by_name(typename);
g_free(typename);
--
1.8.3.2
Re: [Qemu-devel] [PATCHv9 0/5] block: add native support for NFS
Am 03.02.2014 um 10:26 hat Peter Lieven geschrieben: This adds v9 of the NFS protocol driver + qemu-iotest adjustments. Thanks, applied to the block branch. Kevin
Re: [Qemu-devel] Improving patch tracking - something like gerrit?
On Mon, Feb 03, 2014 at 12:45:31PM +, Mark Cave-Ayland wrote: Hi all, It should be fairly evident to most people that the volume of patches flowing through the qemu-devel mailing list is continually increasing, and it is becoming increasingly difficult to track which patches have been applied over time. This is particularly a problem where patchsets have dependencies on other patchsets which haven't yet been applied to git master, which can then cause merge conflicts due to length of time taken for the final series to be merged. Is it time for QEMU to start looking at tools such as gerrit to help manage this process? There seems to be an increasing number of ping requests for outstanding patches (including my own) which don't get applied for weeks, and often even months because they target less popular platforms/subsystems and so don't always get the attention of the committers. Having had to use Gerrit for a long time on OpenStack, I'd never willingly use it on a project I was in charge of for a number of reasons - No practical integration with email based workflows for people who don't want to use web UIs to comment. You can download patches from tool using to view the code outside the UI, but to actually comment you need to use the RSI-inducing, pointy-clicky web UI. - Poor handling of patch series - it shows dependancies between patches but that is basically all it does, and even that has poor UI. People frequently review 1 single patch never noticing that its part of a patch series. There's no way to get a view of all patches in a series ordered correctly. If you tag them with a topic, you can view all patches in the topic, but it randomly re-orders the patches, making it basically useless. - Poor UI for browsing through historical comments on previous versions of the patch. The comments are split between multiple web page views so you again have pointy-clicky hell trying to read through historical comments. - Poor UI for browsing/querying pending patches. Reviewers typically find themselves having to write external/command line tools to query gerrit in order to workaround its limited UI. So sure, gerrit can track every single patch submitted and tell you if it is applied or not, but having used it, I can't say that it is a net win overall, particularly if your development process is heavily using large patch series. Regards, Daniel -- |: http://berrange.com -o-http://www.flickr.com/photos/dberrange/ :| |: http://libvirt.org -o- http://virt-manager.org :| |: http://autobuild.org -o- http://search.cpan.org/~danberr/ :| |: http://entangle-photo.org -o- http://live.gnome.org/gtk-vnc :|
[Qemu-devel] [PATCH 1/3] glib: move g_poll() replacement into glib-compat.h
We have a dedicated header file for wrappers to smooth over glib version
differences. Move the g_poll() definition into glib-compat.h for
consistency.
Signed-off-by: Stefan Hajnoczi stefa...@redhat.com
---
include/glib-compat.h | 12
include/qemu-common.h | 12
2 files changed, 12 insertions(+), 12 deletions(-)
diff --git a/include/glib-compat.h b/include/glib-compat.h
index 8aa77af..8d25900 100644
--- a/include/glib-compat.h
+++ b/include/glib-compat.h
@@ -24,4 +24,16 @@ static inline guint g_timeout_add_seconds(guint interval,
GSourceFunc function,
}
#endif
+#if !GLIB_CHECK_VERSION(2, 20, 0)
+/*
+ * Glib before 2.20.0 doesn't implement g_poll, so wrap it to compile properly
+ * on older systems.
+ */
+static inline gint g_poll(GPollFD *fds, guint nfds, gint timeout)
+{
+GMainContext *ctx = g_main_context_default();
+return g_main_context_get_poll_func(ctx)(fds, nfds, timeout);
+}
+#endif
+
#endif
diff --git a/include/qemu-common.h b/include/qemu-common.h
index 5054836..e0e03cc 100644
--- a/include/qemu-common.h
+++ b/include/qemu-common.h
@@ -124,18 +124,6 @@ int qemu_main(int argc, char **argv, char **envp);
void qemu_get_timedate(struct tm *tm, int offset);
int qemu_timedate_diff(struct tm *tm);
-#if !GLIB_CHECK_VERSION(2, 20, 0)
-/*
- * Glib before 2.20.0 doesn't implement g_poll, so wrap it to compile properly
- * on older systems.
- */
-static inline gint g_poll(GPollFD *fds, guint nfds, gint timeout)
-{
-GMainContext *ctx = g_main_context_default();
-return g_main_context_get_poll_func(ctx)(fds, nfds, timeout);
-}
-#endif
-
/**
* is_help_option:
* @s: string to test
--
1.8.5.3
[Qemu-devel] [PATCH 0/3] glib: move compat functions into glib-compat.h
glib has deprecated APIs like GStaticMutex, g_thread_create(), and others. In QEMU support both old and new APIs since using deprecated APIs would flood us with warnings but legacy distros must continue to build the QEMU source code. This patch series reduces ifdefs by moving glib compat functions into glib-compat.h, where they can be reused. There are two strategies for compat functions: 1. Implement the new API using the deprecated API. This compat function is used when building on a legacy host. Sometimes the API semantics are so different that this option is not feasible. 2. Add a new wrapper API that maps to the deprecated API. The wrapper is not marked deprecated so it works as a drop-in replacement but is implemented using the new API where possible. Stefan Hajnoczi (3): glib: move g_poll() replacement into glib-compat.h glib: add g_thread_new() compat function glib: add compat wrapper for GStaticMutex coroutine-gthread.c | 26 ++ include/glib-compat.h | 44 include/qemu-common.h | 12 trace/simple.c| 31 ++- 4 files changed, 64 insertions(+), 49 deletions(-) -- 1.8.5.3
[Qemu-devel] [PATCH 3/3] glib: add compat wrapper for GStaticMutex
Avoid duplicating ifdefs for the deprecated GStaticMutex API by
introducing compat_g_static_mutex_*() in glib-compat.h.
GStaticMutex users should use the compat API to avoid dealing with
deprecation.
Signed-off-by: Stefan Hajnoczi stefa...@redhat.com
---
coroutine-gthread.c | 13 +++--
include/glib-compat.h | 19 +++
trace/simple.c| 26 +-
3 files changed, 35 insertions(+), 23 deletions(-)
diff --git a/coroutine-gthread.c b/coroutine-gthread.c
index 695c113..28272f8 100644
--- a/coroutine-gthread.c
+++ b/coroutine-gthread.c
@@ -30,7 +30,7 @@ typedef struct {
CoroutineAction action;
} CoroutineGThread;
-static GStaticMutex coroutine_lock = G_STATIC_MUTEX_INIT;
+DEFINE_COMPAT_G_STATIC_MUTEX(coroutine_lock);
/* GLib 2.31 and beyond deprecated various parts of the thread API,
* but the new interfaces are not available in older GLib versions
@@ -123,15 +123,16 @@ static void __attribute__((constructor))
coroutine_init(void)
static void coroutine_wait_runnable_locked(CoroutineGThread *co)
{
while (!co-runnable) {
-g_cond_wait(coroutine_cond, g_static_mutex_get_mutex(coroutine_lock));
+g_cond_wait(coroutine_cond,
+compat_g_static_mutex_get_mutex(coroutine_lock));
}
}
static void coroutine_wait_runnable(CoroutineGThread *co)
{
-g_static_mutex_lock(coroutine_lock);
+compat_g_static_mutex_lock(coroutine_lock);
coroutine_wait_runnable_locked(co);
-g_static_mutex_unlock(coroutine_lock);
+compat_g_static_mutex_unlock(coroutine_lock);
}
static gpointer coroutine_thread(gpointer opaque)
@@ -173,7 +174,7 @@ CoroutineAction qemu_coroutine_switch(Coroutine *from_,
CoroutineGThread *from = DO_UPCAST(CoroutineGThread, base, from_);
CoroutineGThread *to = DO_UPCAST(CoroutineGThread, base, to_);
-g_static_mutex_lock(coroutine_lock);
+compat_g_static_mutex_lock(coroutine_lock);
from-runnable = false;
from-action = action;
to-runnable = true;
@@ -183,7 +184,7 @@ CoroutineAction qemu_coroutine_switch(Coroutine *from_,
if (action != COROUTINE_TERMINATE) {
coroutine_wait_runnable_locked(from);
}
-g_static_mutex_unlock(coroutine_lock);
+compat_g_static_mutex_unlock(coroutine_lock);
return from-action;
}
diff --git a/include/glib-compat.h b/include/glib-compat.h
index ea965df..48fa95c 100644
--- a/include/glib-compat.h
+++ b/include/glib-compat.h
@@ -49,4 +49,23 @@ static inline GThread *g_thread_new(const gchar *unused,
}
#endif
+/* GStaticMutex was deprecated in 2.32.0. Use the compat_g_static_mutex_*()
+ * wrappers to hide the GStaticMutex/GMutex distinction.
+ */
+#if GLIB_CHECK_VERSION(2, 32, 0)
+#define DEFINE_COMPAT_G_STATIC_MUTEX(name) \
+static GMutex name
+
+#define compat_g_static_mutex_lock(m) g_mutex_lock(m)
+#define compat_g_static_mutex_unlock(m) g_mutex_unlock(m)
+#define compat_g_static_mutex_get_mutex(m) (m)
+#else
+#define DEFINE_COMPAT_G_STATIC_MUTEX(name) \
+static GStaticMutex name = G_STATIC_MUTEX_INIT
+
+#define compat_g_static_mutex_lock(m) g_static_mutex_lock(m)
+#define compat_g_static_mutex_unlock(m) g_static_mutex_unlock(m)
+#define compat_g_static_mutex_get_mutex(m) g_static_mutex_get_mutex(m)
+#endif
+
#endif
diff --git a/trace/simple.c b/trace/simple.c
index 8e83e59..dc58afe 100644
--- a/trace/simple.c
+++ b/trace/simple.c
@@ -41,17 +41,7 @@
* Trace records are written out by a dedicated thread. The thread waits for
* records to become available, writes them out, and then waits again.
*/
-#if GLIB_CHECK_VERSION(2, 32, 0)
-static GMutex trace_lock;
-#define lock_trace_lock() g_mutex_lock(trace_lock)
-#define unlock_trace_lock() g_mutex_unlock(trace_lock)
-#define get_trace_lock_mutex() (trace_lock)
-#else
-static GStaticMutex trace_lock = G_STATIC_MUTEX_INIT;
-#define lock_trace_lock() g_static_mutex_lock(trace_lock)
-#define unlock_trace_lock() g_static_mutex_unlock(trace_lock)
-#define get_trace_lock_mutex() g_static_mutex_get_mutex(trace_lock)
-#endif
+DEFINE_COMPAT_G_STATIC_MUTEX(trace_lock);
/* g_cond_new() was deprecated in glib 2.31 but we still need to support it */
#if GLIB_CHECK_VERSION(2, 31, 0)
@@ -151,26 +141,28 @@ static bool get_trace_record(unsigned int idx,
TraceRecord **recordptr)
*/
static void flush_trace_file(bool wait)
{
-lock_trace_lock();
+compat_g_static_mutex_lock(trace_lock);
trace_available = true;
g_cond_signal(trace_available_cond);
if (wait) {
-g_cond_wait(trace_empty_cond, get_trace_lock_mutex());
+g_cond_wait(trace_empty_cond,
+compat_g_static_mutex_get_mutex(trace_lock));
}
-unlock_trace_lock();
+compat_g_static_mutex_unlock(trace_lock);
}
static void wait_for_trace_records_available(void)
{
-lock_trace_lock();
+compat_g_static_mutex_lock(trace_lock);
while (!(trace_available trace_writeout_enabled)) {
[Qemu-devel] [PATCH 2/3] glib: add g_thread_new() compat function
Implement g_thread_new() in terms of the deprecated g_thread_create().
The API was changed in glib 2.31.0.
The compat function allows us to write modern code and avoid ifdefs.
Signed-off-by: Stefan Hajnoczi stefa...@redhat.com
---
coroutine-gthread.c | 13 +++--
include/glib-compat.h | 13 +
trace/simple.c| 5 +
3 files changed, 17 insertions(+), 14 deletions(-)
diff --git a/coroutine-gthread.c b/coroutine-gthread.c
index d3e5b99..695c113 100644
--- a/coroutine-gthread.c
+++ b/coroutine-gthread.c
@@ -76,11 +76,6 @@ static inline void set_coroutine_key(CoroutineGThread *co,
g_private_replace(coroutine_key, co);
}
-static inline GThread *create_thread(GThreadFunc func, gpointer data)
-{
-return g_thread_new(coroutine, func, data);
-}
-
#else
/* Handle older GLib versions */
@@ -104,15 +99,13 @@ static inline void set_coroutine_key(CoroutineGThread *co,
free_on_thread_exit ? (GDestroyNotify)g_free : NULL);
}
+#endif
+
static inline GThread *create_thread(GThreadFunc func, gpointer data)
{
-return g_thread_create_full(func, data, 0, TRUE, TRUE,
-G_THREAD_PRIORITY_NORMAL, NULL);
+return g_thread_new(coroutine, func, data);
}
-#endif
-
-
static void __attribute__((constructor)) coroutine_init(void)
{
if (!g_thread_supported()) {
diff --git a/include/glib-compat.h b/include/glib-compat.h
index 8d25900..ea965df 100644
--- a/include/glib-compat.h
+++ b/include/glib-compat.h
@@ -36,4 +36,17 @@ static inline gint g_poll(GPollFD *fds, guint nfds, gint
timeout)
}
#endif
+#if !GLIB_CHECK_VERSION(2, 31, 0)
+static inline GThread *g_thread_new(const gchar *unused,
+GThreadFunc func,
+gpointer data)
+{
+GThread *thread = g_thread_create(func, data, TRUE, NULL);
+if (!thread) {
+g_error(g_thread_create failed);
+}
+return thread;
+}
+#endif
+
#endif
diff --git a/trace/simple.c b/trace/simple.c
index 57572c4..8e83e59 100644
--- a/trace/simple.c
+++ b/trace/simple.c
@@ -17,6 +17,7 @@
#include pthread.h
#endif
#include qemu/timer.h
+#include glib-compat.h
#include trace.h
#include trace/control.h
#include trace/simple.h
@@ -397,11 +398,7 @@ static GThread *trace_thread_create(GThreadFunc fn)
pthread_sigmask(SIG_SETMASK, set, oldset);
#endif
-#if GLIB_CHECK_VERSION(2, 31, 0)
thread = g_thread_new(trace-thread, fn, NULL);
-#else
-thread = g_thread_create(fn, NULL, FALSE, NULL);
-#endif
#ifndef _WIN32
pthread_sigmask(SIG_SETMASK, oldset, NULL);
--
1.8.5.3
Re: [Qemu-devel] [PATCH v4 0/5] pc: make ACPI pcihp more reusable
On Mon, Feb 03, 2014 at 11:44:56AM +0100, Igor Mammedov wrote: changes since v3: - keep acpi-pci-hotplug-with-bridge-support property and pass its value to acpi_pcihp_init() to init internal .legacy_piix field. changes since v2: - rename 'use_acpi_pci_hotplug' field to 'legacy_piix' and corresponding properties - drop excessive checks for legacy mode - rework legacy vmstate handling to use AcpiPciHpPciStatus[0] structure which reduced acpi/piix4.c by another 50 LOC - move legacy initialization to pcihp.c changes since v1: - add BSEL property to root bus when in compatibility mode as result dropped hw:acpi:pcihp: assume root PCI bus if bus has no ACPI_PCIHP_PROP_BSEL property - rebasing on top of pcihp: reduce number of device check events added patch 3/4. Looks good to me. Applied, thanks a lot. Reuses new pcihp code for handling legacy PCI hotplug in acpi/piix4_pm, which significantly reduces code duplication between piix4_pm and pcihp reducing codebase by ~200 LOC. It also allows cleaner rebase of generic hotplug refactoring, by not increasing code duplication since it will need to be done only for pcihp. Git tree for testing based on mst's PCI tree: https://github.com/imammedo/qemu/commits/pcihp_cleanup_v4 Igor Mammedov (5): pcihp: replace enable|disable_device() with oneliners pcihp: make PCI hotplug mmio handlers indifferent to PCI_HOTPLUG_ADDR pcihp: make pci_read() mmio calback compatible with legacy ACPI hotplug pcihp: remove unused AcpiPciHpPciStatus.device_present field hw:piix4:acpi: reuse pcihp code for legacy PCI hotplug hw/acpi/pcihp.c | 68 --- hw/acpi/piix4.c | 210 --- include/hw/acpi/pcihp.h |5 +- 3 files changed, 55 insertions(+), 228 deletions(-)
[Qemu-devel] [PATCH] block: Fail gracefully with missing filename
This fixes a regression introduced in commit 2a05cbe42 ('block: Allow
block devices without files'):
$ qemu-system-x86_64 -drive driver=file
qemu-system-x86_64: block.c:892: bdrv_open_common: Assertion
`!drv-bdrv_needs_filename || filename != ((void *)0)' failed.
Now the respective check must be performed not only in bdrv_file_open(),
but also in bdrv_open().
Signed-off-by: Kevin Wolf kw...@redhat.com
---
block.c| 11 ++-
tests/qemu-iotests/051 | 12
tests/qemu-iotests/051.out | 24
3 files changed, 42 insertions(+), 5 deletions(-)
diff --git a/block.c b/block.c
index cb21a5f..ac0ccac 100644
--- a/block.c
+++ b/block.c
@@ -832,6 +832,12 @@ static int bdrv_open_common(BlockDriverState *bs,
BlockDriverState *file,
filename = qdict_get_try_str(options, filename);
}
+if (drv-bdrv_needs_filename !filename) {
+error_setg(errp, The '%s' block driver requires a file name,
+ drv-format_name);
+return -EINVAL;
+}
+
trace_bdrv_open_common(bs, filename ?: , flags, drv-format_name);
node_name = qdict_get_try_str(options, node-name);
@@ -1031,11 +1037,6 @@ int bdrv_file_open(BlockDriverState **pbs, const char
*filename,
goto fail;
}
qdict_del(options, filename);
-} else if (drv-bdrv_needs_filename !filename) {
-error_setg(errp, The '%s' block driver requires a file name,
- drv-format_name);
-ret = -EINVAL;
-goto fail;
}
if (!drv-bdrv_file_open) {
diff --git a/tests/qemu-iotests/051 b/tests/qemu-iotests/051
index b23d91b..46345fb 100755
--- a/tests/qemu-iotests/051
+++ b/tests/qemu-iotests/051
@@ -172,6 +172,18 @@ run_qemu -drive file=$TEST_IMG,file.driver=file
run_qemu -drive file=$TEST_IMG,file.driver=qcow2
echo
+echo === Leaving out required options ===
+echo
+
+run_qemu -drive driver=file
+run_qemu -drive driver=nbd
+run_qemu -drive driver=raw
+run_qemu -drive file.driver=file
+run_qemu -drive file.driver=nbd
+run_qemu -drive file.driver=raw
+run_qemu -drive foo=bar
+
+echo
echo === Parsing protocol from file name ===
echo
diff --git a/tests/qemu-iotests/051.out b/tests/qemu-iotests/051.out
index d0c5173..30e2dbd 100644
--- a/tests/qemu-iotests/051.out
+++ b/tests/qemu-iotests/051.out
@@ -225,6 +225,30 @@ Testing: -drive file=TEST_DIR/t.qcow2,file.driver=qcow2
QEMU_PROG: -drive file=TEST_DIR/t.qcow2,file.driver=qcow2: could not open disk
image TEST_DIR/t.qcow2: Block format 'qcow2' used by device '' doesn't support
the option 'filename'
+=== Leaving out required options ===
+
+Testing: -drive driver=file
+QEMU_PROG: -drive driver=file: could not open disk image ide0-hd0: The 'file'
block driver requires a file name
+
+Testing: -drive driver=nbd
+QEMU_PROG: -drive driver=nbd: could not open disk image ide0-hd0: Could not
open image: Invalid argument
+
+Testing: -drive driver=raw
+QEMU_PROG: -drive driver=raw: could not open disk image ide0-hd0: Can't use
'raw' as a block driver for the protocol level
+
+Testing: -drive file.driver=file
+QEMU_PROG: -drive file.driver=file: could not open disk image ide0-hd0: The
'file' block driver requires a file name
+
+Testing: -drive file.driver=nbd
+QEMU_PROG: -drive file.driver=nbd: could not open disk image ide0-hd0: Could
not open image: Invalid argument
+
+Testing: -drive file.driver=raw
+QEMU_PROG: -drive file.driver=raw: could not open disk image ide0-hd0: Can't
use 'raw' as a block driver for the protocol level
+
+Testing: -drive foo=bar
+QEMU_PROG: -drive foo=bar: could not open disk image ide0-hd0: Must specify
either driver or file
+
+
=== Parsing protocol from file name ===
Testing: -hda foo:bar
--
1.8.1.4
Re: [Qemu-devel] [PATCH v2] cris: Add any as alias for crisv32 in user emulation
Am 03.02.2014 14:23, schrieb Edgar E. Iglesias: From: Edgar E. Iglesias edgar.igles...@xilinx.com Signed-off-by: Edgar E. Iglesias edgar.igles...@xilinx.com --- Notes (cl): v1 - v2: Implement as an alias rather than as QOM type. (AF) Reviewed-by: Andreas Färber afaer...@suse.de Thanks, Andreas -- SUSE LINUX Products GmbH, Maxfeldstr. 5, 90409 Nürnberg, Germany GF: Jeff Hawn, Jennifer Guild, Felix Imendörffer; HRB 16746 AG Nürnberg
Re: [Qemu-devel] [PATCH 7/8] roms: update vgabios binaries
Il 05/01/2014 12:45, Jan Kiszka ha scritto: On 2014-01-05 12:26, Paolo Bonzini wrote: Il 05/01/2014 10:35, Jan Kiszka ha scritto: On 2013-12-06 10:35, Gerd Hoffmann wrote: This also switches from lgplvgabios to seavgabios. Signed-off-by: Gerd Hoffmann kra...@redhat.com --- pc-bios/vgabios-cirrus.bin | Bin 35840 - 36864 bytes pc-bios/vgabios-qxl.bin| Bin 40448 - 37376 bytes pc-bios/vgabios-stdvga.bin | Bin 40448 - 37376 bytes pc-bios/vgabios-vmware.bin | Bin 40448 - 37376 bytes pc-bios/vgabios.bin| Bin 40448 - 36864 bytes 5 files changed, 0 insertions(+), 0 deletions(-) This update breaks a 64-bit win7 guest for me. It now spins infinitely in the Starting Windows screen. I'm using stdvga. Is this a known issue? Yes, and IIRC it's fixed in seabios.git. (Hmm, cd roms; make seavgabios doesn't work with out-of-tree builds) Confirmed, win7 boots fine again with current seabios.git master. So please update qemu. Ping? Paolo
Re: [Qemu-devel] [PATCH 7/8] roms: update vgabios binaries
Hi, Confirmed, win7 boots fine again with current seabios.git master. So please update qemu. Ping? Oops, totally forgot that. I'll send out a pull req in a moment. cheers, Gerd
Re: [Qemu-devel] [PATCH 00/12] trace: [tcg] Allow tracing guest events in TCG-generated code
On Fri, Jan 31, 2014 at 05:09:03PM +0100, Lluís Vilanova wrote: Adds the base ability to specify which events in the trace-events file may be used to trace guest activity in the TCG code (using the tcg event propery). Such events generate an extra set of tracing functions that can be called during TCG code generation and will automatically redirect a call to the appropriate backend-dependent tracing functions when the guest code is executed. I've never worked on TCG but this seems like a good idea: a way to plant trace events in TCG-generated code. CCing TCG folks to check they are happy with the approach Once any high-level discussion is done I'll review the tracing changes. Stefan
[Qemu-devel] [PULL 0/3] Update seabios to 1.7.4
Hi, Sorry for the delay folks, totally forgot that one. Here is the seabios update to 1.7.4 final. please pull, Gerd The following changes since commit 2f61120c10da9128357510debc8e66880cd2bfdc: Merge remote-tracking branch 'qmp-unstable/queue/qmp' into staging (2014-02-01 23:32:31 +) are available in the git repository at: git://git.kraxel.org/qemu tags/pull-roms-1 for you to fetch changes up to 41419b0f11d125ad792660638eb452c767eddc28: Update seabios binaries to 1.7.4 (2014-02-03 15:40:01 +0100) Update seabios to 1.7.4 Bruce Rogers (1): roms: remove explicit MAKEFLAGS from recursive make invocations Gerd Hoffmann (2): Update seabios submodule to 1.7.4 Update seabios binaries to 1.7.4 pc-bios/bios-256k.bin | Bin 262144 - 262144 bytes pc-bios/bios.bin | Bin 131072 - 131072 bytes pc-bios/vgabios-cirrus.bin | Bin 36864 - 36864 bytes pc-bios/vgabios-qxl.bin| Bin 37376 - 37376 bytes pc-bios/vgabios-stdvga.bin | Bin 37376 - 37376 bytes pc-bios/vgabios-vmware.bin | Bin 37376 - 37376 bytes pc-bios/vgabios.bin| Bin 36864 - 36864 bytes roms/Makefile | 22 +++--- roms/seabios | 2 +- 9 files changed, 12 insertions(+), 12 deletions(-)
