Re: Uploading linux (3.12.5-1)
On Tue, December 17, 2013 6:27 pm, Ben Hutchings wrote: I intend to upload linux version 3.12.5-1 to sid on Tuesday or Wednesday. If there are any important changes that should be included (particularly if there are known regressions), please commit them or let me know about them. commit 5cfe8f1ba5eebe6f4b6e5858cdb1a5be4f3272a6 alpha: Prevent a NULL ptr dereference in csum_partial_copy. does not appear to have made it to the 3.12 stable queue and without it the 3.12.x kernel is likely to crash on Alpha. Thanks Michael. -- To UNSUBSCRIBE, email to debian-kernel-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org Archive: http://lists.debian.org/60369.121.74.15.242.1387270260.squir...@mail.orcon.net.nz
Bug#685894: linux FTBFS on alpha: gcc-4.5 no longer available
Source: linux Version: 3.2.23-1 Severity: important User: debian-al...@lists.debian.org Usertags: alpha Tags: patch X-Debbugs-CC: debian-al...@lists.debian.org Justification: Fails to build from source but built in the past. linux no longer builds on Alpha due to the demise of gcc-4.5 in unstable. We will have to move to the default compiler (gcc-4.6) but using gcc-4.6 to build the kernel results in relocation errors against the sbss section during the link. This can be worked around by modifying the -msmall-data compiler argument in arch/alpha/Makefile to -mlarge-data. It means extra CPU instructions to access kernel global variables thus a slightly larger and less efficient kernel. I did try to ask the gcc Alpha maintainer why using gcc-4.6 should have such an effect but did not get back a particularly informative answer. So if you could remove the requirement to build the kernel with gcc-4.5 and apply the attached patch to switch to the large data model for the next linux source upload, I should be grateful. Cheers Michael. diff --git a/arch/alpha/Makefile b/arch/alpha/Makefile index 4759fe7..b5d44bd 100644 --- a/arch/alpha/Makefile +++ b/arch/alpha/Makefile @@ -12,7 +12,7 @@ NM := $(NM) -B LDFLAGS_vmlinux := -static -N #-relax CHECKFLAGS += -D__alpha__ -m64 -cflags-y := -pipe -mno-fp-regs -ffixed-8 -msmall-data +cflags-y := -pipe -mno-fp-regs -ffixed-8 -mlarge-data cflags-y += $(call cc-option, -fno-jump-tables) cpuflags-$(CONFIG_ALPHA_EV4) := -mcpu=ev4
Bug#664834: linux-tools: please build perf utils on Alpha.
Package: linux-tools Version: 3.2.7-2 Severity: wishlist User: debian-al...@lists.debian.org Usertags: alpha X-Debbugs-CC: debian-al...@lists.debian.org Could we have the perf tools built on Alpha please. For some reason they are not in the linux-tools package, whereas they are for other arches. Performance hardware event support for EV67/68 CPUs is in the kernel and it would be nice to be able to use the perf tools on those CPUs. FWIW, I did do a make within the tools/perf directory of the source package and the tools built successfully. Thanks, Michael. -- To UNSUBSCRIBE, email to debian-kernel-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org Archive: http://lists.debian.org/4f69820b.8040...@orcon.net.nz
Bug#658460: SOCK_NONBLOCK doubly defined in system headers on Alpha
Package: linux-libc-dev
Version: 3.2.1-1
Severity: normal
User: debian-al...@lists.debian.org
Usertags: alpha
X-Debbugs-CC: debian-al...@lists.debian.org
The simple test program:
#include linux/net.h
#include sys/socket.h
fails to compile on the Alpha arch with the following:
$ gcc -c test.c
In file included from /usr/include/alpha-linux-gnu/sys/socket.h:40:0,
from test.c:2:
/usr/include/alpha-linux-gnu/bits/socket.h:68:3: error: expected
identifier before numeric constant
It is because linux/net.h eventually includes
/usr/include/alpha-linux-gnu/asm/socket.h which ends with the following:
/* O_NONBLOCK clashes with the bits used for socket types. Therefore we
* have to define SOCK_NONBLOCK to a different value here.
*/
#define SOCK_NONBLOCK 0x4000
Then sys/socket.h includes /usr/include/alpha-linux-gnu/bits/socket.h
(from libc6.1-dev) which has the following:
enum __socket_type
{
...snip...
SOCK_NONBLOCK = 0x4000/* Atomically mark descriptor(s) as
non-blocking. */
}
which with the above macro definition becomes
0x4000 = 0x4000
inside the enumeration, and the compiler understandingly barfs. Indeed
this is the cause of the FTBFS of the package audit which compiles on
all other arches without a problem but fails on Alpha due to the double
definition of SOCK_NONBLOCK.
I suspect that SOCK_NONBLOCK should not be defined in asm/socket.h in
linux-libc-dev.
Cheers
Michael.
--
To UNSUBSCRIBE, email to debian-kernel-requ...@lists.debian.org
with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org
Archive: http://lists.debian.org/4f2baa7d.8080...@orcon.net.nz
Bug#657112: Please build linux-2.6 on Alpha with gcc-4.5
Source: linux-2.6 Version: 3.2.1-1 Severity: important User: debian-al...@lists.debian.org Usertags: alpha X-Debbugs-CC: debian-al...@lists.debian.org Justification: Crashes system in flaming heap of process/CPU lockups. gcc-4.4 on Alpha produces bad code and is the reason for many crashes that existed in quite of number of packages. The bug is fixed in gcc-4.5. This is relevant to the kernel because the source package linux-2.6 requests that gcc-4.4 is used to build the kernel. Ouch! I hadn't noticed that until now, when, wow, we finally (!) get a relatively new Debian kernel built on Alpha, so I installed it with gusto and, strike a light, it took my flash new buildd down in a flaming heap. See my earlier report on this problem to the linux kernel list at: https://lkml.org/lkml/2012/1/19/410 Please update the version of gcc-4.4 used to build the kernel on Alpha to gcc-4.5. I suggest using gcc-4.5 rather than the default gcc-4.6 as some people have reported that gcc-4.6 extends the data segment beyond 64kb (arch/alpha/Makefile specifies -msmall-data so the data segment cannot exceed 64kB) thus linking fails. Actually, on a Debian kernel that might not be a problem since almost everything is built as a module, but we desperately need a _working_ kernel on Alpha so I suggest not risking gcc-4.6 at the moment. Cheers Michael. -- To UNSUBSCRIBE, email to debian-kernel-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org Archive: http://lists.debian.org/4f1e51fc.6090...@orcon.net.nz
Bug#649468: [PATCH] alpha: add io{read,write}{16,32}be functions
These functions are used in some PCI drivers with big-endian MMIO space. Admittedly it is almost certain that no one this side of the Moon would use such a card in an Alpha but it does get us closer to being able to build allyesconfig or allmodconfig, and it enables the Debian default generic config to build. Signed-off-by: Michael Cree mc...@orcon.net.nz --- arch/alpha/include/asm/io.h |5 + 1 files changed, 5 insertions(+), 0 deletions(-) diff --git a/arch/alpha/include/asm/io.h b/arch/alpha/include/asm/io.h index 56ff965..6365ef2 100644 --- a/arch/alpha/include/asm/io.h +++ b/arch/alpha/include/asm/io.h @@ -490,6 +490,11 @@ extern inline void writeq(u64 b, volatile void __iomem *addr) } #endif +#define ioread16be(p) be16_to_cpu(ioread16(p)) +#define ioread32be(p) be32_to_cpu(ioread32(p)) +#define iowrite16be(v,p) iowrite16(cpu_to_be16(v), (p)) +#define iowrite32be(v,p) iowrite32(cpu_to_be32(v), (p)) + #define inb_p inb #define inw_p inw #define inl_p inl -- 1.7.4.3 -- To UNSUBSCRIBE, email to debian-kernel-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org Archive: http://lists.debian.org/1322640100-14534-1-git-send-email-mc...@orcon.net.nz
Bug#649468: FTBFS on Alpha: undefined iowrite32be wanted by module mlx4_en.
Source: linux-2.6 Version: 3.1.1-1 Severity: Important User: debian-al...@lists.debian.org Usertags: alpha X-Debbugs-CC: debian-al...@lists.debian.org The package linux-2,6 FTBFS on Alpha. From the log: Building modules, stage 2. MODPOST 2591 modules ERROR: iowrite32be [drivers/net/mlx4/mlx4_en.ko] undefined! Full log is available at: http://buildd.debian-ports.org/status/fetch.php?pkg=linux-2.6arch=alphaver=3.1.1-1stamp=1321281816 In an experiment I changed the config to remove the module mlx4_en from being built and it successfully built to completion. Cheers Michael. -- To UNSUBSCRIBE, email to debian-kernel-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org Archive: http://lists.debian.org/4eca04f4.80...@orcon.net.nz
Bug#649468: FTBFS on Alpha: undefined iowrite32be wanted by module mlx4_en.
On 22/11/2011, at 3:14 AM, Ben Hutchings wrote:
On Mon, 2011-11-21 at 20:59 +1300, Michael Cree wrote:
The package linux-2,6 FTBFS on Alpha. From the log:
Building modules, stage 2.
MODPOST 2591 modules
ERROR: iowrite32be [drivers/net/mlx4/mlx4_en.ko] undefined!
[...]
This needs to be added to arch/alpha similarly to:
commit 06901bd83412db5a31de7526e637101ed0c2c472
Author: Arnd Bergmann a...@arndb.de
Date: Sat Sep 3 17:54:44 2011 +0200
ARM: add io{read,write}{16,32}be functions
Thanks, I'll prepare a patch for that. It is likely to be later this
week since I am busy the next couple of days.
Cheers
Michael.
--
To UNSUBSCRIBE, email to debian-kernel-requ...@lists.debian.org
with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org
Archive:
http://lists.debian.org/8beb1eb1-b31b-46a3-9ea6-ad1795bfc...@orcon.net.nz
Bug#647586: FTBFS on alpha: missing module crc32
Source: linux-2.6 Version: 3.0.0-6 Severity: Important The package linux-2.6 FTBFS on alpha. From the log: kernel-wedge install-files 3.0.0-2 install -D -m 644 debian/linux-image-3.0.0-2-alpha-generic/boot/vmlinuz-3.0.0-2-alpha-generic debian/kernel-image-3.0.0-2-alpha-generic-di/boot/vmlinuz install -D -m 644 debian/linux-image-3.0.0-2-alpha-generic/boot/System.map-3.0.0-2-alpha-generic debian/kernel-image-3.0.0-2-alpha-generic-di/boot/System.map kernel-wedge copy-modules 3.0.0-2 alpha-generic 3.0.0-2-alpha-generic missing module crc32 command exited with status 1 The full log is available at: http://buildd.debian-ports.org/status/fetch.php?pkg=linux-2.6arch=alphaver=3.0.0-6stamp=1320290779 Cheers Michael. -- To UNSUBSCRIBE, email to debian-kernel-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org Archive: http://lists.debian.org/4eb3a62c.3090...@orcon.net.nz
Bug#647587: Alpha: missing accept4 syscall
Source: linux-2.6 Version: 3.0.0-6 Severity: Important User: debian-al...@lists.debian.org Usertags: alpha X-Debbugs-CC: debian-al...@lists.debian.org The kernel for the alpha port is missing the accept4 syscall; the implementation, long overdue, is now in Linus' kernel as commit 0a8c384ebe4cc1c462076732dbffb595e7e81d5c. Could we please have this cherry-picked to the 3.0.0 kernel (or the 3.1.0 kernel if that is what you plan to upload to unstable next). It should apply without modification. The new version of udev in unstable depends on this syscall and without it it floods the system logs with error messages filling the disk in short order. Cheers Michael. -- To UNSUBSCRIBE, email to debian-kernel-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org Archive: http://lists.debian.org/4eb3ac0d.1060...@orcon.net.nz
Bug#647587: Alpha: missing accept4 syscall
On 05/11/11 04:07, Ben Hutchings wrote: On Fri, 2011-11-04 at 22:10 +1300, Michael Cree wrote: The kernel for the alpha port is missing the accept4 syscall; the implementation, long overdue, is now in Linus' kernel as commit 0a8c384ebe4cc1c462076732dbffb595e7e81d5c. Could we please have this cherry-picked to the 3.0.0 kernel (or the 3.1.0 kernel if that is what you plan to upload to unstable next). It should apply without modification. Done for 3.1; I don't think we'll do another 3.0 upload. Thanks. Kernel 3.1 is cool --- it means we (alpha) will get working KMS on Radeon graphics too! The new version of udev in unstable depends on this syscall and without it it floods the system logs with error messages filling the disk in short order. Was there a stable release for alpha based on squeeze? If so, that should also be updated. No. :-( Cheers Michael. -- To UNSUBSCRIBE, email to debian-kernel-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org Archive: http://lists.debian.org/4eb437c9.1020...@orcon.net.nz
Bug#647586: FTBFS on alpha: missing module crc32
On 05/11/11 02:29, Ben Hutchings wrote: On Fri, 2011-11-04 at 21:45 +1300, Michael Cree wrote: The package linux-2.6 FTBFS on alpha. From the log: kernel-wedge install-files 3.0.0-2 install -D -m 644 debian/linux-image-3.0.0-2-alpha-generic/boot/vmlinuz-3.0.0-2-alpha-generic debian/kernel-image-3.0.0-2-alpha-generic-di/boot/vmlinuz install -D -m 644 debian/linux-image-3.0.0-2-alpha-generic/boot/System.map-3.0.0-2-alpha-generic debian/kernel-image-3.0.0-2-alpha-generic-di/boot/System.map kernel-wedge copy-modules 3.0.0-2 alpha-generic 3.0.0-2-alpha-generic missing module crc32 command exited with status 1 I'm not going to spend any time on this. It has to be fixed by the alpha porters. The kernel-wedge configuration is the same as was previously used for alpha, so it was apparently already broken. The Alpha installer has had no work done on it since Lenny I suspect. I see that the file debian/installer/alpha/modules/alpha/crc-modules lists crc32. I can't find any crc32 module in the kernel. Maybe the crc-modules file should be modified to be: #include crc-modules as it is on some of the other architectures. Or as you suggest we could disable building udebs on alpha for the present and leave this as a problem to fix for the future. Cheers Michael. -- To UNSUBSCRIBE, email to debian-kernel-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org Archive: http://lists.debian.org/4eb442d6.3090...@orcon.net.nz
Bug#638696: Processed: Bug#638696: linux-2.6 still FTBFS on alpha
On 28/08/11 22:40, Ben Hutchings wrote:
On Sun, 2011-08-28 at 10:03 +, Debian Bug Tracking System wrote:
Processing commands for cont...@bugs.debian.org:
reopen 638696
Bug #638696 {Done: Ben Hutchings b...@decadent.org.uk} [linux-2.6] FTBFS
on alpha: module sht15 needs gpiolib not selected by config.
'reopen' may be inappropriate when a bug has been closed with a version;
you may need to use 'found' to remove fixed versions.
As I suspected, there are other problematic drivers. It looks like
GENERIC_GPIO should really only be enabled by GPIOLIB. Try this:
Applied on clean upstream 3.0.3 source and used the config from an alpha
generic build. Compiles fine with an Alpha cross-compiler with
GENERIC_GPIO set to no.
Cheers
Michael.
--
To UNSUBSCRIBE, email to debian-kernel-requ...@lists.debian.org
with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org
Archive: http://lists.debian.org/4e5aa024.2070...@orcon.net.nz
Bug#638696: FTBFS on alpha: module sht15 needs gpiolib not selected by config.
Package: linux-2.6 Version: 3.0.0-2 Severity: normal From the build log: CC [M] drivers/hwmon/sht15.o In file included from +/build/buildd-linux-2.6_3.0.0-2-alpha-uWIvrQ/linux-2.6-3.0.0/debian/build/source_alpha_none/arch/alp +ha/include/asm/gpio.h:20, from +/build/buildd-linux-2.6_3.0.0-2-alpha-uWIvrQ/linux-2.6-3.0.0/debian/build/source_alpha_none/include/ +linux/gpio.h:18, from +/build/buildd-linux-2.6_3.0.0-2-alpha-uWIvrQ/linux-2.6-3.0.0/debian/build/source_alpha_none/drivers/ +hwmon/sht15.c:21: /build/buildd-linux-2.6_3.0.0-2-alpha-uWIvrQ/linux-2.6-3.0.0/debian/build/source_alpha_none/include/a +sm-generic/gpio.h: In function 'gpio_get_value_cansleep': /build/buildd-linux-2.6_3.0.0-2-alpha-uWIvrQ/linux-2.6-3.0.0/debian/build/source_alpha_none/include/a +sm-generic/gpio.h:223: error: implicit declaration of function 'gpio_get_value' Full build log available at: http://buildd.debian-ports.org/status/fetch.php?pkg=linux-2.6arch=alphaver=3.0.0-2stamp=1313880281 Setting CONFIG_GPIOLIB should fix this. Regards Michael Cree. -- To UNSUBSCRIBE, email to debian-kernel-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org Archive: http://lists.debian.org/4e509eaa.6000...@orcon.net.nz
Bug#589804: Implement missing syscalls on Alpha
Package: linux-2.6 Version: 2.6.32-17 Severity: wishlist Tags: patch It would be nice to have the missing syscalls on Alpha wired up to bring it up to date with the other architectures. It was unfortunate that the patch sent upstream to achieve this was missed during the 2.6.32 development window and didn't get merged by Linus unto 2.6.33. I attach the patch patch-alpha-wire-up-syscalls.txt that wires up the missing syscalls on the Alpha architecture and brings it up to date with the other architectures. This patch is a modification of the upstream commits 6e17e8b9fb74b9fb9f6ea331f7f4a049c5b4c4b8 and 21797c599c710d3851d241c4b50690f2482bf618 in that it replaces the recvmmsg syscall with a placeholder as it is not available in 2.6.32. As mentioned in bug report #588409 (which is now closed) support for performance events on Alpha was merged upstream for 2.6.33. I include here the patch patch-alpha-implement-sw-perf-events.txt which is a combination of the upstream commits a582e6f01b90211933e70edcec9bc0bbb1157402 and fcd14b3203b538dca04a2b065c774c0b57863eec and wires up the performance events syscall and provides the Alpha specific code to the perf tools. I have tested both these patches applied to the Debian 2.6.32-17 kernel source (though I did have to disable the buildcheck.py script as the ABI didn't match) and it is running successfully on an Alpha PWS600au. Michael. -- Package-specific info: ** Kernel log: boot messages should be attached ** Model information system type : Tsunami system variation: Monet system revision : 0 platform string : COMPAQ Professional Workstation XP1000 ** PCI devices: :00:07.0 ISA bridge [0601]: Contaq Microsystems 82c693 [1080:c693] Control: I/O+ Mem+ BusMaster+ SpecCycle- MemWINV- VGASnoop- ParErr- Stepping- SERR- FastB2B- DisINTx- Status: Cap- 66MHz- UDF- FastB2B+ ParErr- DEVSEL=medium TAbort- TAbort- MAbort- SERR- PERR- INTx- Latency: 0 :00:07.1 IDE interface [0101]: Contaq Microsystems 82c693 [1080:c693] (prog-if 80 [Master]) Control: I/O+ Mem- BusMaster+ SpecCycle- MemWINV- VGASnoop- ParErr- Stepping- SERR- FastB2B- DisINTx- Status: Cap- 66MHz- UDF- FastB2B+ ParErr- DEVSEL=medium TAbort- TAbort- MAbort- SERR- PERR- INTx- Latency: 0 Interrupt: pin A routed to IRQ 14 Region 0: I/O ports at 01f0 [size=8] Region 1: I/O ports at 03f4 [size=1] Region 4: I/O ports at 9440 [size=16] Kernel driver in use: pata_cypress :00:07.2 IDE interface [0101]: Contaq Microsystems 82c693 [1080:c693] (prog-if 00 []) Control: I/O+ Mem- BusMaster+ SpecCycle- MemWINV- VGASnoop- ParErr- Stepping- SERR- FastB2B- DisINTx- Status: Cap- 66MHz- UDF- FastB2B+ ParErr- DEVSEL=medium TAbort- TAbort- MAbort- SERR- PERR- INTx- Latency: 0 Interrupt: pin B routed to IRQ 15 Region 0: I/O ports at 0170 [size=8] Region 1: I/O ports at 0374 [size=1] Region 4: Memory at 0930 (32-bit, non-prefetchable) [disabled] [size=64K] :00:07.3 USB Controller [0c03]: Contaq Microsystems 82c693 [1080:c693] (prog-if 10 [OHCI]) Control: I/O+ Mem+ BusMaster+ SpecCycle- MemWINV- VGASnoop- ParErr- Stepping- SERR- FastB2B- DisINTx- Status: Cap- 66MHz- UDF- FastB2B+ ParErr- DEVSEL=medium TAbort- TAbort- MAbort- SERR- PERR- INTx- Latency: 248, Cache Line Size: 32 bytes Interrupt: pin A routed to IRQ 10 Region 0: Memory at 0931 (32-bit, non-prefetchable) [size=4K] Kernel driver in use: ohci_hcd :00:0b.0 PCI bridge [0604]: PLX Technology, Inc. PEX8112 x1 Lane PCI Express-to-PCI Bridge [10b5:8112] (rev aa) (prog-if 00 [Normal decode]) Control: I/O+ Mem+ BusMaster+ SpecCycle- MemWINV- VGASnoop- ParErr- Stepping- SERR- FastB2B- DisINTx- Status: Cap+ 66MHz+ UDF- FastB2B- ParErr- DEVSEL=medium TAbort- TAbort- MAbort- SERR- PERR- INTx- Latency: 255, Cache Line Size: 64 bytes Bus: primary=00, secondary=01, subordinate=01, sec-latency=255 I/O behind bridge: 8000-8fff Memory behind bridge: 0900-090f Prefetchable memory behind bridge: 1000-27ff Secondary status: 66MHz- FastB2B- ParErr- DEVSEL=fast TAbort- TAbort- MAbort+ SERR- PERR- BridgeCtl: Parity+ SERR+ NoISA+ VGA+ MAbort- Reset- FastB2B- PriDiscTmr- SecDiscTmr- DiscTmrStat- DiscTmrSERREn- Capabilities: access denied :00:0c.0 USB Controller [0c03]: VIA Technologies, Inc. VT82x UHCI USB 1.1 Controller [1106:3038] (rev 62) (prog-if 00 [UHCI]) Subsystem: VIA Technologies, Inc. VT82x UHCI USB 1.1 Controller [1106:3038] Control: I/O+ Mem+ BusMaster+ SpecCycle- MemWINV- VGASnoop- ParErr- Stepping- SERR- FastB2B- DisINTx- Status: Cap+ 66MHz- UDF- FastB2B- ParErr- DEVSEL=medium TAbort- TAbort- MAbort- SERR- PERR- INTx- Latency: 255, Cache Line Size: 64
Bug#588409: perf (linux-tools-2.6.32) fails to build on many architectures
Regarding the perf tools: Commit 5d7bdab75cd56d2bdc0986ae5546be3b09fea70a upstream should give the -fstack-protector-all test. Regarding the Alpha architecture the following commits that are in v2.6.33 of the kernel upstream provide the performance event interface: Commit fcd14b3203b538dca04a2b065c774c0b57863eec upstream adds the Alpha specific part for perf tools. Commit a582e6f01b90211933e70edcec9bc0bbb1157402 upstream provides minimal support for performance events in the kernel for Alpha. However it probably won't apply cleanly to 2.6.32 without commit 21797c599c710d3851d241c4b50690f2482bf618 which wires up quite a number of missing syscalls, which in turn depends on 6e17e8b9fb74b9fb9f6ea331f7f4a049c5b4c4b8 which wires up the syscall for recvmmsg that is new in 2.6.33. Thus a modified 21797c599 is required for 2.6.32 to do all this; I think I might still have the patch somewhere, but in any case I could regenerate it should you wish to have it. Cheers Michael -- To UNSUBSCRIBE, email to debian-kernel-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org Archive: http://lists.debian.org/4c35b72d.5000...@orcon.net.nz
Bug#583949: 2.6.32-3-amd64 can't read DVDs/CDs (sometimes) - Bug#583949
On 23/06/10 00:22, zoltan herman wrote: Take a look at the 551101 bug report, Ben H. wrote Date: Mon, 21 Jun 2010 11:11:32 +0100 We switched from IDE drivers to libata-based drivers in 2.6.32-10. My suspicion is that this is a bug in ide-cd which will no longer apply. In my case that is not the reason for the errors. It appears that my DVD errors are due to an interaction between two problems: a bug that was temporarily in mplayer SVN source, and not setting the region code on the DVD drive. Having now updated mplayer and set the region code on my DVD drive I haven't seen the DVD drive errors again, but I haven't done a lot of testing either. I wonder whether this is the explanation (or not) for the problems the original reporter of the bug was having? Cheers Michael. -- To UNSUBSCRIBE, email to debian-kernel-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org Archive: http://lists.debian.org/4c2114d9.6000...@orcon.net.nz
Bug#583949: 2.6.32-3-amd64 can't read DVDs/CDs (sometimes)
I am seeing this bug too on both AMD64 and Alpha architectures. Sometimes DVD plays up with error messages (similar to those given by original reporter; see below) and doesn't seem to recover until a reboot. When the errors start typically I can get the disc mounted (unlike the original reporter) but can't read much from it. My hardware is as follows: AMD64 machine running Debian testing: kernel: 2.6.32-3-amd64 (Debian kernel) SATA controller: 00:09.0 IDE interface: nVidia Corporation MCP78S [GeForce 8200] SATA Controller (non-AHCI mode) (rev a2) DVD drive: Lite-On ATAPI iHAS120 6 7L0F I tried an old unstable installation (it hasn't been upgraded for 2 or 3 months) that is still hanging around on another partition. It has the 2.6.32-trunk-amd64 Debian kernel and it seemed to work okay, but I really need to test it further to be really sure as I have since discovered that the disc errors may take a few disc changes before they start to appear. Alpha running Debian unstable: kernel: 2.6.34 (compiled from www.kernel.org) SATA controller: 0001:03:09.0 RAID bus controller: Silicon Image, Inc. SiI 3114 [SATALink/SATARaid] Serial ATA Controller (rev 02) DVD drive: Lite-On ATAPI DVD A DH20A3S 9P54 On the Alpha I took it back to a 2.6.32.7 kernel and still could get the disc errors to appear. To take the kernel back any earlier I will probably have to create a patched kernel (udev didn't create the necessary system disk device nodes when I tried an even earlier kernel -- missing system calls maybe?). This Alpha also has an old Compaq IDE CD-ROM drive which I use for booting. It has also taken to spitting out errors since about 2.6.32 kernel or maybe 2.6.31, but I have really never cared since I only use it for booting and that is working fine. Example of the DVD errors I am getting on the Alpha: [240094.915157] UDF-fs: Partition marked readonly; forcing readonly mount [240094.960079] UDF-fs INFO UDF: Mounting volume 'SPARTACUS', timestamp 2000/11/02 17:15 (1000) [240099.128045] sr 2:0:0:0: [sr0] Result: hostbyte=DID_OK driverbyte=DRIVER_SENSE [240099.128045] sr 2:0:0:0: [sr0] Sense Key : Illegal Request [current] [240099.128045] sr 2:0:0:0: [sr0] Add. Sense: Read of scrambled sector without authentication [240099.128045] sr 2:0:0:0: [sr0] CDB: Read(10): 28 00 00 00 26 f4 00 00 40 00 [240099.128045] end_request: I/O error, dev sr0, sector 39888 [240099.128045] quiet_error: 97 callbacks suppressed [240099.128045] Buffer I/O error on device sr0, logical block 9972 [240099.128045] Buffer I/O error on device sr0, logical block 9973 [240099.128045] Buffer I/O error on device sr0, logical block 9974 [240099.128045] Buffer I/O error on device sr0, logical block 9975 [240099.128045] Buffer I/O error on device sr0, logical block 9976 [240099.128045] Buffer I/O error on device sr0, logical block 9977 [240099.129022] Buffer I/O error on device sr0, logical block 9978 [240099.129022] Buffer I/O error on device sr0, logical block 9979 [240099.129022] Buffer I/O error on device sr0, logical block 9980 [240099.129022] Buffer I/O error on device sr0, logical block 9981 [240099.160272] sr 2:0:0:0: [sr0] Result: hostbyte=DID_OK driverbyte=DRIVER_SENSE [240099.160272] sr 2:0:0:0: [sr0] Sense Key : Illegal Request [current] [240099.160272] sr 2:0:0:0: [sr0] Add. Sense: Read of scrambled sector without authentication [240099.160272] sr 2:0:0:0: [sr0] CDB: Read(10): 28 00 00 00 27 34 00 00 20 00 [240099.160272] end_request: I/O error, dev sr0, sector 40144 Cheers Michael. -- To UNSUBSCRIBE, email to debian-kernel-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org Archive: http://lists.debian.org/4c0b7803.4040...@orcon.net.nz
Bug#527265: Strange QLA1040 firmaware load error with 2.6.30 from testing
Uwe Schindler wrote: After upgrading to linux-image-2.6.30-alpha-generic I get the following trace in the kernel log. The qlogic firmware package is installed, as initramfs complained about missing bin files. The driver loads and the system works as expected, all disks are reported: [4.416013] qla1280 :00:09.0: firmware: requesting qlogic/1040.bin [4.416013] [ cut here ] [4.416013] WARNING: at /build/buildd-linux-2.6_2.6.30-5-alpha-u828kA/linux-2 .6-2.6.30/debian/build/source_alpha_none/kernel/softirq.c:141 _local_bh_enable_i Anybody has seen this, too? Yes. I normally run my own kernel from kernel.org but the Debian kernel has finally caught up to the leading edge and I thought I'd give it a go. With the Debian kernel I also see the warning and backtrace. The 2.6.30.5 kernel I compiled from kernel.org source does not give the warning. That might be because I select the kernel config option to include the firmwire in the kernel binary. Just one question: Is the firmware always needed or can I uninstall the qlogic-firmware package again? Does the controller has an eprom or does it only works with driver-provided firmware? As the firmware does not seem to be loaded correctly, is it really needed? When I first tried the Debian 2.6.30 kernel the kernel locked up when initialising the qlogic driver as I had completely forgotten about the firmware. After installing qlogic-firmware driver it then booted fine. Given I was running a kernel that included the firmware prior to running the Debian kernel that is suggestive that the firmware is reloaded on every boot. Cheers Michael. -- To UNSUBSCRIBE, email to debian-kernel-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org
Bug#473877: linux-image-2.6.24-1-alpha-generic: fail to boot
Kernels since 2.6.23 (and its release candidates) break with an older version of aboot. If you have been upgrading your system from Etch or a version of Lenny from before about middle of last year you will need to upgrade the aboot package to the latest, and you may need to manually reinstall the boot block with the swriteboot command. Be aware that if you get this operation wrong you screw up your computer and won't be able to boot at all, unless you have another disc in the system that is bootable. See http://thread.gmane.org/gmane.linux.debian.ports.alpha/7247/focus=7267 for a hint on using swriteboot and to jump midway into the thread discussing the aboot issue. Michael. -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Bug#476766: linux-image-2.6.24-1-alpha-generic: Alpha using sound driver causes kernel OOPS
Package: linux-image-2.6.24-1-alpha-generic Version: 2.6.24-5 Severity: critical Justification: breaks the whole system Playing a sound file via alsa sound devices for the second time causes mmio related kernel OOPS on certain Alpha architectures. Observed on miata (PWS600au) while playing via onboard es1888/7 sound chips and on Tsunami (XP1000) when playing sound through certain sound cards (but onboard es1888 seems to work okay!). Rated the bug as critical as it causes kernel OOPS and often takes out the complete system. This issue is fixed in kernel 2.6.25. Patch is available as commit c143d43aa3149b83e4b40624a27aa2b18638afec in Linus' git tree. Would be nice to have this patch applied if 2.6.24 is going to be released in Etch+half and Lenny. -- System Information: Debian Release: lenny/sid APT prefers testing APT policy: (500, 'testing') Architecture: alpha Kernel: Linux 2.6.24.4 Locale: LANG=en_NZ, LC_CTYPE=en_NZ (charmap=ISO-8859-1) Shell: /bin/sh linked to /bin/bash Versions of packages linux-image-2.6.24-1-alpha-generic depends on: ii debconf [debconf-2.0] 1.5.20 Debian configuration management sy ii initramfs-tools [linux-initra 0.91e tools for generating an initramfs ii module-init-tools 3.4-1 tools for managing Linux kernel mo linux-image-2.6.24-1-alpha-generic recommends no packages. -- debconf information excluded -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
