The original patch missed the last empty line in target-unicore32/helper.c
So I re-format it.
Original patch From: Aurelien Jarno aurel...@aurel32.net
This patch moves FPU helpers to op_helper.c, which has direct access to
the env pointer. This means one less argument to pass, so a gain in
Applied, thanks.
Tested-and-Acked-by: Guan Xuetao g...@mprc.pku.edu.cn
Guan
-Original Message-
From: Aurelien Jarno [mailto:aurel...@aurel32.net]
Sent: Wednesday, April 13, 2011 6:31 AM
To: qemu-devel@nongnu.org
Cc: Aurelien Jarno; Guan Xuetao
Subject: [PATCH 3/3] target-unicore32:
---
.gitignore |3 +++
Makefile |5 -
2 files changed, 7 insertions(+), 1 deletions(-)
diff --git a/.gitignore b/.gitignore
index 1d79680..08013fc 100644
--- a/.gitignore
+++ b/.gitignore
@@ -44,6 +44,9 @@ QMP/qmp-commands.txt
*.ky
*.log
*.pdf
+*.cps
+*.fns
+*.kys
*.pg
*.pyc
We ported MIPS64 r2 user mode emulation. When a binary is given to
qemu-mips64, our code first check whether it is Octeon binary or not if yes
it enable Octeon specific Instructions for. The following code snippet do
this job:
diff --git a/linux-user/elfload.c b/linux-user/elfload.c
index
This series finally fixes -icount with iothread and avoids deadlocks
due to the vm_clock not making progress when the VM is stopped.
The crux of the fix is in patch 1, while patch 2 implements the
clock warping that fixes deadlocks in v2. Clock warping uses
the nanosecond resolution rt_clock
The previous patch however is not enough, because if the virtual CPU
goes to sleep waiting for a future timer interrupt to wake it up, qemu
deadlocks. The timer interrupt never comes because time is driven by
icount, but the vCPU doesn't run any insns.
You could say that VCPUs should never go to
It is purely for icount-based virtual timers. And now that we got the
code right, rename the function to clarify the intended scope.
Signed-off-by: Paolo Bonzini pbonz...@redhat.com
Tested-by: Edgar E. Iglesias edgar.igles...@gmail.com
---
cpus.c |4 ++--
qemu-timer.c | 13
The correct fix for -icount is to consider the biggest difference
between iothread and non-iothread modes. In the traditional model,
CPUs run _before_ the iothread calls select (or WaitForMultipleObjects
for Win32). In the iothread model, CPUs run while the iothread
isn't holding the mutex, i.e.
This reverts commits 225d02cd and c9f7383c. While some parts of
the latter could be saved, I preferred a smooth, complete revert.
Signed-off-by: Paolo Bonzini pbonz...@redhat.com
Tested-by: Edgar E. Iglesias edgar.igles...@gmail.com
---
qemu-timer.c | 66
Use get_option_parameter() to instead of duplicating the loop, and use
BDRV_SECTOR_SIZE to instead of 512
Signed-off-by: Mitnick Lyu mitnick@gmail.com
---
block/vpc.c |8 ++--
1 files changed, 2 insertions(+), 6 deletions(-)
diff --git a/block/vpc.c b/block/vpc.c
index
1 fixes spellos in the mass-storage driver
2-4 fix issues in Linux usb pass-through code.
Signed-off-by: Brad Hards br...@frogmouth.net
---
hw/usb-msd.c |4 ++--
1 files changed, 2 insertions(+), 2 deletions(-)
diff --git a/hw/usb-msd.c b/hw/usb-msd.c
index 76f5b02..947fd3f 100644
--- a/hw/usb-msd.c
+++ b/hw/usb-msd.c
@@ -33,7 +33,7 @@ do { printf(usb-msd: fmt , ## __VA_ARGS__);
This isn't used, but leaving it empty causes valgrind noise.
Signed-off-by: Brad Hards br...@frogmouth.net
---
usb-linux.c |1 +
1 files changed, 1 insertions(+), 0 deletions(-)
diff --git a/usb-linux.c b/usb-linux.c
index 255009f..d958853 100644
--- a/usb-linux.c
+++ b/usb-linux.c
@@
Signed-off-by: Brad Hards br...@frogmouth.net
---
usb-linux.c |4 ++--
1 files changed, 2 insertions(+), 2 deletions(-)
diff --git a/usb-linux.c b/usb-linux.c
index d958853..1f33c2c 100644
--- a/usb-linux.c
+++ b/usb-linux.c
@@ -107,7 +107,7 @@ enum {
/*
* Control transfer state.
* Note
Signed-off-by: Brad Hards br...@frogmouth.net
---
usb-linux.c | 16
1 files changed, 8 insertions(+), 8 deletions(-)
diff --git a/usb-linux.c b/usb-linux.c
index 1f33c2c..b02a0f9 100644
--- a/usb-linux.c
+++ b/usb-linux.c
@@ -233,8 +233,8 @@ static void async_complete(void
Am 12.04.2011 18:06, schrieb Amit Shah:
The GET_EVENT_STATUS_NOTIFICATION ATAPI command is listed as a
mandatory command in the spec but we don't really implement it any of
its sub-commands.
The commit message for the last commit explains why implementing just
the media subcommand is
Am 13.04.2011 11:30, schrieb Mitnick Lyu:
Use get_option_parameter() to instead of duplicating the loop, and use
BDRV_SECTOR_SIZE to instead of 512
Signed-off-by: Mitnick Lyu mitnick@gmail.com
Thanks, applied to the block branch.
Kevin
On Tue, 12 Apr 2011, Jan Kiszka wrote:
Well, either you have a use for the VCPU state (how do you do migration
in Xen without it?), or you should probably teach QEMU in a careful
clean way to run its device model without VCPUs - and without any
TCG-related memory consumption. For the latter,
On 2011-04-13 12:56, Stefano Stabellini wrote:
On Tue, 12 Apr 2011, Jan Kiszka wrote:
Well, either you have a use for the VCPU state (how do you do migration
in Xen without it?), or you should probably teach QEMU in a careful
clean way to run its device model without VCPUs - and without any
Signed-off-by: Alon Levy al...@redhat.com
---
libcacard/vcard_emul_nss.c |4 ++--
1 files changed, 2 insertions(+), 2 deletions(-)
diff --git a/libcacard/vcard_emul_nss.c b/libcacard/vcard_emul_nss.c
index 71f2ba3..baada52 100644
--- a/libcacard/vcard_emul_nss.c
+++
On Wed, 13 Apr 2011, Jan Kiszka wrote:
On 2011-04-13 12:56, Stefano Stabellini wrote:
On Tue, 12 Apr 2011, Jan Kiszka wrote:
Well, either you have a use for the VCPU state (how do you do migration
in Xen without it?), or you should probably teach QEMU in a careful
clean way to run its
Were you able to replicate the problem with the steps that I had mentioned ?
The key thing is is if you don't set breakpoint at 0x4 or 0x8 just follow
the execution flow using si command of GDB.
You will definitely hit the problem.
Ah, I had to find a different gdb to reproduce this with
The following changes since commit 9df38c47d01eb1fd7eb9d60ac70a4170e638b4a2:
target-arm: Detect tininess before rounding for FP operations (2011-04-12
23:33:33 +0200)
are available in the git repository at:
git://repo.or.cz/qemu/kevin.git for-anthony
Amit Shah (7):
atapi: Drives can
From: Stefan Hajnoczi stefa...@linux.vnet.ibm.com
Zero data clusters are a space-efficient way of storing zeroed regions
of the image.
Signed-off-by: Stefan Hajnoczi stefa...@linux.vnet.ibm.com
Signed-off-by: Kevin Wolf kw...@redhat.com
---
docs/specs/qed_spec.txt |8
1 files
From: Amit Shah amit.s...@redhat.com
Table 629 of the MMC-5 spec mentions two different error conditions when
a CDROM eject is requested: a) while a disc is inserted and b) while a
disc is not inserted.
Ensure we return the appropriate error for the present condition of the
drive and disc
From: Anthony Liguori aligu...@us.ibm.com
Zero clusters are similar to unallocated clusters except instead of reading
their value from a backing file when one is available, the cluster is always
read as zero.
This implements read support only. At this stage, QED will never write a
zero cluster.
From: Amit Shah amit.s...@redhat.com
Drivers are free to lock drives without any media present. Such a
condition should not result in an error condition.
See Table 341 in MMC-5 spec for details.
Signed-off-by: Amit Shah amit.s...@redhat.com
Signed-off-by: Kevin Wolf kw...@redhat.com
---
From: Amit Shah amit.s...@redhat.com
Handle GET_EVENT_STATUS_NOTIFICATION's No Event Available response in a
generic way so that future additions to the code to handle other
response types is easier.
Signed-off-by: Amit Shah amit.s...@redhat.com
Acked-by: Jes Sorensen jes.soren...@redhat.com
From: Mitnick Lyu mitnick@gmail.com
Use get_option_parameter() to instead of duplicating the loop, and
use BDRV_SECTOR_SIZE to instead of 512
Signed-off-by: Mitnick Lyu mitnick@gmail.com
Signed-off-by: Kevin Wolf kw...@redhat.com
---
block/vpc.c |8 ++--
1 files changed, 2
From: Amit Shah amit.s...@redhat.com
After a media change, the only commands allowed from the guest were
REQUEST_SENSE and INQUIRY. The guest may also issue
GET_EVENT_STATUS_NOTIFICATION commands to get media
changed notification.
Signed-off-by: Amit Shah amit.s...@redhat.com
Acked-by: Jes
From: Amit Shah amit.s...@redhat.com
Instead of using magic numbers, use structs that are more descriptive of
the fields being used.
Signed-off-by: Amit Shah amit.s...@redhat.com
Acked-by: Jes Sorensen jes.soren...@redhat.com
Acked-by: Paolo Bonzini pbonz...@redhat.com
Signed-off-by: Kevin Wolf
From: Amit Shah amit.s...@redhat.com
Implement the 'media' sub-command of the GET_EVENT_STATUS_NOTIFICATION
command. This helps us report tray open, tray closed, no media, media
present states to the guest.
Newer Linux kernels (2.6.38+) rely on this command to revalidate discs
after media
From: Amit Shah amit.s...@redhat.com
This makes the code more readable.
Also, there's a block like:
if () {
...
} else {
...
}
Split that into
if () {
...
return;
}
...
Signed-off-by: Amit Shah amit.s...@redhat.com
Acked-by: Jes Sorensen jes.soren...@redhat.com
Acked-by: Paolo
On Wed, Apr 13, 2011 at 10:45 AM, Brad Hards br...@frogmouth.net wrote:
Signed-off-by: Brad Hards br...@frogmouth.net
---
usb-linux.c | 16
1 files changed, 8 insertions(+), 8 deletions(-)
Some of these printfs look like they might be useful given that the
current USB
On Wed, Apr 13, 2011 at 10:45 AM, Brad Hards br...@frogmouth.net wrote:
1 fixes spellos in the mass-storage driver
2-4 fix issues in Linux usb pass-through code.
I left a comment on patch 4 and will wait for activity on that email
thread. Otherwise these could go via the trivial-patches tree
Hello Kevin, Stefan and all
There is a sentence The maximum size of a dynamic hard disk is 2040 GB.
The actual size is restricted by the underlying disk hardware protocol. For
example,
ATA hard disks have a 127-GB limit in VHD specification. This limitation
infected the
algorithm of
Hi,
On 04/13/2011 11:45 AM, Brad Hards wrote:
Signed-off-by: Brad Hardsbr...@frogmouth.net
---
usb-linux.c | 16
1 files changed, 8 insertions(+), 8 deletions(-)
diff --git a/usb-linux.c b/usb-linux.c
index 1f33c2c..b02a0f9 100644
--- a/usb-linux.c
+++ b/usb-linux.c
@@
This is the result of an of the list discussion with
me, and I can confirm this fixes the issues I was seeing:
Tested-by: Hans de Goede hdego...@redhat.com
On 04/13/2011 01:42 PM, Alon Levy wrote:
Signed-off-by: Alon Levyal...@redhat.com
---
libcacard/vcard_emul_nss.c |4 ++--
1 files
Am 13.04.2011 14:47, schrieb Lyu Mitnick:
Hello Kevin, Stefan and all
There is a sentence The maximum size of a dynamic hard disk is 2040 GB.
The actual size is restricted by the underlying disk hardware protocol.
For example,
ATA hard disks have a 127-GB limit in VHD specification. This
On Wed, Apr 13, 2011 at 1:43 PM, Anthony Liguori aligu...@us.ibm.com wrote:
diff --git a/roms/seabios b/roms/seabios
index cc97564..06d0bdd 16
--- a/roms/seabios
+++ b/roms/seabios
@@ -1 +1 @@
-Subproject commit cc975646af69f279396d4d5e1379ac6af80ee637
+Subproject commit
On 2011-04-13 13:49, Stefano Stabellini wrote:
On Wed, 13 Apr 2011, Jan Kiszka wrote:
On 2011-04-13 12:56, Stefano Stabellini wrote:
On Tue, 12 Apr 2011, Jan Kiszka wrote:
Well, either you have a use for the VCPU state (how do you do migration
in Xen without it?), or you should probably teach
On Tue, 12 Apr 2011 21:31:18 +0300
Blue Swirl blauwir...@gmail.com wrote:
On Tue, Apr 12, 2011 at 10:52 AM, Avi Kivity a...@redhat.com wrote:
On 04/11/2011 08:15 PM, Blue Swirl wrote:
On Mon, Apr 11, 2011 at 10:01 AM, Markus Armbrusterarm...@redhat.com
wrote:
Avi
On 04/13/2011 08:04 AM, Stefan Hajnoczi wrote:
On Wed, Apr 13, 2011 at 1:43 PM, Anthony Liguorialigu...@us.ibm.com wrote:
diff --git a/roms/seabios b/roms/seabios
index cc97564..06d0bdd 16
--- a/roms/seabios
+++ b/roms/seabios
@@ -1 +1 @@
-Subproject commit
From: Anthony Liguori aligu...@us.ibm.com
QEMU can drop a backing file so that an image file no longer depends on
the backing file, but this feature has not been exposed in qemu-img.
This is useful in an image streaming usecase or when an image file has
been fully allocated and no reads can hit
On 12 April 2011 22:59, Aurelien Jarno aurel...@aurel32.net wrote:
Use float64_unordered() in helper_cmptun() instead of doing the
the comparison manually. This also fixes the wrong behaviours with
sNaNs.
Signed-off-by: Aurelien Jarno aurel...@aurel32.net
---
target-alpha/op_helper.c |
On 12 April 2011 22:59, Aurelien Jarno aurel...@aurel32.net wrote:
On alpha, all NaN should trap during a comparison, not only sNaN. Fix
this by using float64_eq() instead of float64_eq_quiet().
The Compiler Writer's Guide disagrees with you:
On Wed, 13 Apr 2011, Jan Kiszka wrote:
On 2011-04-13 13:49, Stefano Stabellini wrote:
On Wed, 13 Apr 2011, Jan Kiszka wrote:
On 2011-04-13 12:56, Stefano Stabellini wrote:
On Tue, 12 Apr 2011, Jan Kiszka wrote:
Well, either you have a use for the VCPU state (how do you do migration
in
Thanks, Jan. I split my patch into three and started a new thread. I also
put all options in -net user. Yes, TCP firewall is coming. You'll see some
of the added functions will be shared.
Dai
On Tue, Apr 12, 2011 at 9:38 AM, Jan Kiszka jan.kis...@siemens.com wrote:
On 2011-04-12 18:19, Daisuke
On 13 April 2011 16:38, Richard Henderson r...@twiddle.net wrote:
[ Odd, the original thread doesn't seem to have arrived here. ]
On 04/13/2011 07:52 AM, Peter Maydell wrote:
So I think you want the _quiet version here. (And helper_cmpteq
needs to use float64_eq_quiet rather than float64_eq.)
On 12 April 2011 22:59, Aurelien Jarno aurel...@aurel32.net wrote:
This patch series started with the goal of improving the build of
target-i386 with softfloat (instead of softfloat-native), but it slowly
became a collection of fixes and improvements with regard to softfloat
and targets FPU.
I think you are right. This seems to be more of a GDB issue.
Any ways thanks for your support.
--Anup
On Wed, Apr 13, 2011 at 5:24 PM, Peter Maydell
peter.mayd...@linaro.orgwrote:
Were you able to replicate the problem with the steps that I had
mentioned ?
The key thing is is if you don't
I think you are right. This seems to be more of a GDB issue.
The debug stub is still part of QEMU, so let's not close this bug just
yet :-)
** Summary changed:
- Undefined instruction exception starts at offset 0x8 instead of 0x4
+ ARM: singlestepping insn which UNDEFs should stop at UNDEF
Signed-off-by: Lluís Vilanova vilan...@ac.upc.edu
---
target-arm/helper.c| 246 ++--
target-arm/iwmmxt_helper.c | 28 +++--
target-arm/translate.c | 26 ++---
3 files changed, 150 insertions(+), 150 deletions(-)
diff --git
Signed-off-by: Lluís Vilanova vilan...@ac.upc.edu
---
target-m68k/translate.c |2 +-
1 files changed, 1 insertions(+), 1 deletions(-)
diff --git a/target-m68k/translate.c b/target-m68k/translate.c
index 038c0af..15a650e 100644
--- a/target-m68k/translate.c
+++ b/target-m68k/translate.c
@@
This provides a consistent naming scheme across all targets.
Signed-off-by: Lluís Vilanova vilan...@ac.upc.edu
---
target-m68k/helper.c|2 +-
target-m68k/helper.h| 54 +++
target-m68k/helpers.h | 54
This provides a consistent naming scheme across all targets.
Signed-off-by: Lluís Vilanova vilan...@ac.upc.edu
---
target-arm/helper.c|2
target-arm/helper.h| 475
target-arm/helpers.h | 475
On 13 April 2011 17:38, Lluís xscr...@gmx.net wrote:
Signed-off-by: Lluís Vilanova vilan...@ac.upc.edu
---
target-arm/helper.c | 246
++--
target-arm/iwmmxt_helper.c | 28 +++--
target-arm/translate.c | 26 ++---
3 files changed, 150
On 04/13/2011 08:42 AM, Peter Maydell wrote:
On 13 April 2011 16:38, Richard Henderson r...@twiddle.net wrote:
[ Odd, the original thread doesn't seem to have arrived here. ]
On 04/13/2011 07:52 AM, Peter Maydell wrote:
So I think you want the _quiet version here. (And helper_cmpteq
needs to
[ Odd, the original thread doesn't seem to have arrived here. ]
On 04/13/2011 07:52 AM, Peter Maydell wrote:
So I think you want the _quiet version here. (And helper_cmpteq
needs to use float64_eq_quiet rather than float64_eq.)
Yes, the _quiet version is what's needed for all comparisons.
For
Public bug reported:
The latest git version of qemu (commit
9df38c47d01eb1fd7eb9d60ac70a4170e638b4a2) fails to boot the OpenSolaris
image from http://dlc.sun.com/osol/opensolaris/2009/06/osol-0906-ai-
sparc.iso.
qemu-img create opensolaris 3G
qemu-system-sparc -hda opensolaris -cdrom
On Tue, Apr 12, 2011 at 11:41:49PM +0100, Peter Maydell wrote:
On 12 April 2011 22:59, Aurelien Jarno aurel...@aurel32.net wrote:
I haven't look at the documentation, but for the neighbouring code it looks
clear that floatx80_eq() should be used instead of floatx80_eq_quiet().
Actually I
On Tue, Apr 12, 2011 at 07:40:33PM -0700, Nathan Froyd wrote:
On Tue, Apr 12, 2011 at 11:59:29PM +0200, Aurelien Jarno wrote:
Given that float32_*() functions are IEEE754 compliant, the efscmp*()
functions are correctly implemented, while efstst*() are not. This
patch reverse the
On Wed, Apr 13, 2011 at 12:26:01PM +0500, Khansa Butt wrote:
We ported MIPS64 r2 user mode emulation. When a binary is given to
qemu-mips64, our code first check whether it is Octeon binary or not if yes
it enable Octeon specific Instructions for. The following code snippet do
this job:
On Wed, Apr 13, 2011 at 08:53:27AM -0700, Richard Henderson wrote:
On 04/13/2011 08:42 AM, Peter Maydell wrote:
On 13 April 2011 16:38, Richard Henderson r...@twiddle.net wrote:
[ Odd, the original thread doesn't seem to have arrived here. ]
On 04/13/2011 07:52 AM, Peter Maydell wrote:
On Wed, Apr 13, 2011 at 06:16:55PM +0100, Peter Maydell wrote:
On 13 April 2011 17:38, Lluís xscr...@gmx.net wrote:
Signed-off-by: Lluís Vilanova vilan...@ac.upc.edu
---
target-arm/helper.c | 246
++--
target-arm/iwmmxt_helper.c | 28
On Wed, Apr 13, 2011 at 4:08 PM, Luiz Capitulino lcapitul...@redhat.com wrote:
On Tue, 12 Apr 2011 21:31:18 +0300
Blue Swirl blauwir...@gmail.com wrote:
On Tue, Apr 12, 2011 at 10:52 AM, Avi Kivity a...@redhat.com wrote:
On 04/11/2011 08:15 PM, Blue Swirl wrote:
On Mon, Apr 11, 2011 at
Aurelien Jarno writes:
On Wed, Apr 13, 2011 at 06:16:55PM +0100, Peter Maydell wrote:
On 13 April 2011 17:38, Lluís xscr...@gmx.net wrote:
Signed-off-by: Lluís Vilanova vilan...@ac.upc.edu
---
target-arm/helper.c | 246
++--
Aurelien Jarno writes:
On Wed, Apr 13, 2011 at 06:16:55PM +0100, Peter Maydell wrote:
On 13 April 2011 17:38, Lluís xscr...@gmx.net wrote:
Signed-off-by: Lluís Vilanova vilan...@ac.upc.edu
---
target-arm/helper.c | 246
++--
On Mon, Apr 11, 2011 at 10:27:41PM +0200, Jan Kiszka wrote:
These patches were posted before. They bring down the overhead of the
io-thread mode for TCG here, specifically when emulating SMP.
The major change in this version, besides rebasing, is the exclusion of
KVM from the main loop
gen_pc_load was introduced in commit
d2856f1ad4c259e5766847c49acbb4e390731bd4.
The only reason for parameter searched_pc was
a debug statement in target-i386/translate.c.
Remove searched_pc from the debug statement
and from the parameter list of gen_pc_load.
Signed-off-by: Stefan Weil
helpfull - helpful
usefull - useful
cotrol - control
and a grammar fix.
Signed-off-by: Stefan Weil w...@mail.berlios.de
---
qemu-options.hx |4 ++--
savevm.c|2 +-
target-arm/helper.c |2 +-
3 files changed, 4 insertions(+), 4 deletions(-)
diff --git
On Tue, Apr 12, 2011 at 7:19 PM, Daisuke Nojiri dnoj...@google.com wrote:
This patch adds: -drop-udp, -allow-udp ADDR:PORT, -drop-log FILE
e.g.) $ qemu -net user -drop-log qemu.drop -drop-udp -allow-udp
10.0.2.3:53
-drop-udp enables usermode firewall for out-going UDP packats from a guest.
Hello Stefan,
I have a question about get_option_parameter(). I am wondering whether
get_option_parameter is suitable to use instead of doing the search by
myself
in the case like following:
/* Read out options */
while (options options-name) {
if (!strcmp(options-name,
On 13 April 2011 21:38, Stefan Weil w...@mail.berlios.de wrote:
gen_pc_load was introduced in commit
d2856f1ad4c259e5766847c49acbb4e390731bd4.
The only reason for parameter searched_pc was
a debug statement in target-i386/translate.c.
Remove searched_pc from the debug statement
and from the
On Tue, Apr 12, 2011 at 11:31:20PM +0100, Peter Maydell wrote:
On 12 April 2011 22:32, Aurelien Jarno aurel...@aurel32.net wrote:
On Mon, Apr 11, 2011 at 04:32:08PM +0100, Peter Maydell wrote:
@@ -1524,12 +1528,12 @@ uint64_t HELPER(neon_abdl_u16)(uint32_t a,
uint32_t b)
{
On Wed, 13 Apr 2011 10:52:37 pm Hans de Goede wrote:
@@ -359,7 +359,7 @@ static int usb_host_claim_interfaces(USBHostDevice
*dev, int configuration)
ret = ioctl(dev-fd, USBDEVFS_CLAIMINTERFACE,interface);
if (ret 0) {
if (errno == EBUSY) {
-
This patch series started with the goal of improving the build of
target-i386 with softfloat (instead of softfloat-native), but it slowly
became a collection of fixes and improvements with regard to softfloat
and targets FPU.
Aurelien Jarno (19):
softfloat: use GCC builtins to count the leading
Add floatx_{add,mul,sub} defines, and use them instead of using direct
C operations.
Reviewed-by: Peter Maydell peter.mayd...@linaro.org
Signed-off-by: Aurelien Jarno aurel...@aurel32.net
---
target-i386/exec.h |6 ++
target-i386/op_helper.c | 18 --
2 files
Use float unions from cpu-all.h instead of redefining new (wrong for arm)
ones in target-i386. This also allows building cpu-exec.o with softfloat.
Reviewed-by: Peter Maydell peter.mayd...@linaro.org
Signed-off-by: Aurelien Jarno aurel...@aurel32.net
---
target-i386/exec.h | 27
Add a CPU_LDoubleU type, matching the floatx80 definition and the long
double type on x86 hosts.
Based on a patch from Laurent Vivier laur...@vivier.eu.
Cc: Laurent Vivier laur...@vivier.eu
Reviewed-by: Peter Maydell peter.mayd...@linaro.org
Signed-off-by: Aurelien Jarno aurel...@aurel32.net
---
SSE instructions CMPUNORDPS/D and CMPORDPS/D do not trigger an invalid
exception if operands are qNANs.
Reviewed-by: Peter Maydell peter.mayd...@linaro.org
Signed-off-by: Aurelien Jarno aurel...@aurel32.net
---
target-i386/ops_sse.h |4 ++--
1 files changed, 2 insertions(+), 2 deletions(-)
Add float*_unordered_quiet() functions to march the softfloat versions.
As FPU status is not tracked with softfloat-native, they don't differ
from the signaling version.
Reviewed-by: Peter Maydell peter.mayd...@linaro.org
Signed-off-by: Aurelien Jarno aurel...@aurel32.net
---
I am not a big fan of code moving, but having the signaling version in
the middle of quiet versions and vice versa doesn't make the code easy
to read.
This patch is a simple code move, basically swapping locations of
float*_eq and float*_eq_quiet.
Reviewed-by: Peter Maydell
As the softfloat comparison functions already test for NaN, there is no
need to always call the float*_unordered*() functions.
Reviewed-by: Peter Maydell peter.mayd...@linaro.org
Signed-off-by: Aurelien Jarno aurel...@aurel32.net
---
target-mips/op_helper.c | 72
MIPS FPU instructions should start with a clean softfpu status. This
is done for the most instructions, but not for comparison ones.
Signed-off-by: Aurelien Jarno aurel...@aurel32.net
---
target-mips/op_helper.c | 41 +
1 files changed, 25 insertions(+),
Softfloat has its own implementation to count the leading zeros. However
a lot of architectures have either a dedicated instruction or an
optimized to do that. When using GCC = 3.4, this patch uses GCC builtins
instead of the handcoded implementation.
Note that I amware that QEMU_GNUC_PREREQ is
Use CPU_LDoubleU in cpu_dump_state() instead of redefining a union for
doing the conversion.
Based on a patch from Laurent Vivier laur...@vivier.eu.
Cc: Laurent Vivier laur...@vivier.eu
Reviewed-by: Peter Maydell peter.mayd...@linaro.org
Signed-off-by: Aurelien Jarno aurel...@aurel32.net
---
Contrary to cabs.ps.* instructions, c.ps.* should not compare the absolute
value of the operand, but directly the operands.
Signed-off-by: Aurelien Jarno aurel...@aurel32.net
---
target-mips/op_helper.c |8
1 files changed, 4 insertions(+), 4 deletions(-)
diff --git
Use float64_unordered_quiet() in helper_cmptun() instead of doing the
the comparison manually.
According to the Alpha Compiler Writer's Guide, we should use the
_quiet version here, as CMPTUN and CMPTEQ should generate InvalidOp
for SNaNs but not for QNaNs.
Thanks to Peter Maydell
Make clear for all comparison functions which ones trigger an exception
for all NaNs, and which one only for sNaNs.
Reviewed-by: Peter Maydell peter.mayd...@linaro.org
Signed-off-by: Aurelien Jarno aurel...@aurel32.net
---
fpu/softfloat.c | 85
Reviewed-by: Peter Maydell peter.mayd...@linaro.org
Signed-off-by: Aurelien Jarno aurel...@aurel32.net
---
target-mips/op_helper.c | 35 ---
1 files changed, 20 insertions(+), 15 deletions(-)
diff --git a/target-mips/op_helper.c b/target-mips/op_helper.c
index
Use the new float*_unordered*() functions from softfloat instead of
redefining a private version.
Reviewed-by: Peter Maydell peter.mayd...@linaro.org
Signed-off-by: Aurelien Jarno aurel...@aurel32.net
---
target-mips/op_helper.c | 168 ---
1 files
float*_eq functions have a different semantics than other comparison
functions. Fix that by first renaming float*_quiet() into float*_eq_quiet().
Note that it is purely mechanical, and the behaviour should be unchanged.
That said it clearly highlight problems due to this different semantics,
they
Add float*_unordered() functions to softfloat, matching the softfloat-native
ones. Also add float*_unordered_quiet() functions to match the others
comparison functions.
This allow target-i386/ops_sse.h to be compiled with softfloat.
Reviewed-by: Peter Maydell peter.mayd...@linaro.org
float*_eq_signaling functions have a different semantics than other
comparison functions. Fix that by renaming float*_quiet_signaling() into
float*_eq().
Note that it is purely mechanical, and the behaviour should be unchanged.
Reviewed-by: Peter Maydell peter.mayd...@linaro.org
Signed-off-by:
efstst*() functions are fast SPE funtions which do not take into account
special values (infinites, NaN, etc.), while efscmp*() functions are
IEEE754 compliant.
Given that float32_*() functions are IEEE754 compliant, the efscmp*()
functions are correctly implemented, while efstst*() are not. This
On Sat, Apr 2, 2011 at 19:36, Stefan Weil w...@mail.berlios.de wrote:
tcg_gen_exit_tb takes a parameter of type tcg_target_long,
so the type casts of pointer to long should be replaced by
type casts of pointer to tcg_target_long (suggested by Blue Swirl).
These changes are needed for build
Ping.
Gleb, The issue of specifying both data and file is addressed.
Do you have any other comments? Otherwise can you provide your acked-by?
On Tue, Apr 05, 2011 at 10:53:16PM +0400, Michael Tokarev wrote:
This patch almost rewrites acpi_table_add() function
(but still leaves it using old
Am 14.04.2011 03:42, schrieb TeLeMan:
On Sat, Apr 2, 2011 at 19:36, Stefan Weil w...@mail.berlios.de wrote:
tcg_gen_exit_tb takes a parameter of type tcg_target_long,
so the type casts of pointer to long should be replaced by
type casts of pointer to tcg_target_long (suggested by Blue Swirl).
99 matches
Mail list logo
