commit dc53fffc105f68cb08ca872acd51550e89aa2e67 Merge: f560902 46b952a Author: Linus Torvalds Date: Mon Jul 6 14:07:00 2009 -0700 Merge branch 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/jbarnes/pci-2.6 * 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/jbarnes/pci-2.6: PCI: Fix IRQ swizzling for ARI-enabled devices ia64/PCI: adjust section annotation for pcibios_setup() x86/PCI: get root CRS before scanning children x86/PCI: fix boundary checking when using root CRS PCI MSI: Fix restoration of MSI/MSI-X mask states in suspend/resume PCI MSI: Unmask MSI if setup failed PCI MSI: shorten PCI_MSIX_ENTRY_* symbol names PCI: make pci_name() take const argument PCI: More PATA quirks for not entering D3 PCI: fix kernel-doc warnings PCI: check if bus has a proper bridge device before triggering SBR PCI: remove pci_dac_dma_... APIs on mn10300 PCI ECRC: Remove unnecessary semicolons PCI MSI: Return if alloc_msi_entry for MSI-X failed commit f560902c2d39c255bd67c0211f5dd80edb97a712 Merge: 9861df1 db6e3f9 Author: Linus Torvalds Date: Mon Jul 6 14:05:57 2009 -0700 Merge branch 'fixes-for-linus' of git://git.monstr.eu/linux-2.6-microblaze * 'fixes-for-linus' of git://git.monstr.eu/linux-2.6-microblaze: microblaze: Fix cast warning for init.c microblaze: Wire up new syscalls microblaze: use generic syscalls.h microblaze: clean up signal handling microblaze: convert all simple headers to use asm-generic microblaze: use the generic lib/checksum.c microblaze: fall back on generic header files for the ABI commit 9861df15f44d98eb6fa9a839b558633ecee87194 Merge: 8871b20 67fc25e Author: Linus Torvalds Date: Mon Jul 6 14:05:09 2009 -0700 Merge branch 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/penberg/slab-2.6 * 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/penberg/slab-2.6: SLAB: Fix lockdep annotations fix RCU-callback-after-kmem_cache_destroy problem in sl[aou]b commit 8871b201da6e443b7ab2967050e3064360d0604f Merge: 085ff82 0a68434 Author: Linus Torvalds Date: Mon Jul 6 14:04:25 2009 -0700 Merge branch 'for_linus' of git://git.kernel.org/pub/scm/linux/kernel/git/mchehab/linux-2.6 * 'for_linus' of git://git.kernel.org/pub/scm/linux/kernel/git/mchehab/linux-2.6: (25 commits) V4L/DVB (12206): get_dvb_firmware: Correct errors in MPC718 firmware extraction logic V4L/DVB (12203): radio-si470x: fix lock imbalance V4L/DVB (12202): em28xx, fix lock imbalance V4L/DVB (12172): em28xx: Add autodetection code for Silvercrest 1.3 mpix V4L/DVB (12171): em28xx: fix webcam usage with different output formats V4L/DVB (12169): em28xx-video: fix VIDIOC_G_FMT and VIDIOC_ENUMFMT with webcams V4L/DVB (12156): em28xx: Fix tuning for Terratec Cinergy T XS USB (zl10353 version) V4L/DVB (12139): em28xx: add other video formats V4L/DVB (12138): em28xx: add support for Silvercrest Webcam V4L/DVB (12174): mt9v011: let's stick with datasheet values where it works V4L/DVB (12173): mt9v011: properly calculate image resolution registers V4L/DVB (12137): mt9v011: CodingStyle fixes V4L/DVB (12136): mt9v011: Some fixes at the register initialization table V4L/DVB (12135): Add a driver for mt9v011 sensor V4L/DVB (12166): cx23885: add FIXME comment above set_frontend override V4L/DVB (12165): cx23885: override set_frontend to allow rf input path switching on the HVR1275 V4L/DVB (12148): move V4L2_PIX_FMT_SGRBG8 to the proper place V4L/DVB (12182): cx18: Add DVB-T support for Yuan MPC-718 cards with an MT352 or ZL10353 V4L/DVB (12181): get_dvb_firmware: Add Yuan MPC718 MT352 DVB-T "firmware" extraction V4L/DVB (12180): cx18: Update Yuan MPC-718 card entry with better information and guesses ... commit 085ff82c9c615d502d1e6754879d8d4ff590d143 Merge: f63bafe5 3dfc813 Author: Linus Torvalds Date: Mon Jul 6 14:03:59 2009 -0700 Merge git://git.infradead.org/iommu-2.6 * git://git.infradead.org/iommu-2.6: intel-iommu: Don't use identity mapping for PCI devices behind bridges intel-iommu: Use iommu_should_identity_map() at startup time too. intel-iommu: No mapping for non-PCI devices intel-iommu: Restore DMAR_BROKEN_GFX_WA option for broken graphics drivers intel-iommu: Add iommu_should_identity_map() function intel-iommu: Fix reattaching of devices to identity mapping domain intel-iommu: Don't set identity mapping for bypassed graphics devices intel-iommu: Fix dma vs. mm page confusion with aligned_nrpages() commit f63bafe55654caf3a62f73500eafd1b89ca6f7ff Merge: 4148df9 ebbb16b Author: Linus Torvalds Date: Mon Jul 6 14:03:44 2009 -0700 Merge branch 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/ieee1394/linux1394-2.6 * 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/ieee1394/linux1394-2.6: ieee1394: sbp2: add support for disks >2 TB (and 16 bytes long CDBs) firewire: sbp2: add support for disks >2 TB (and 16 bytes long CDBs) firewire: core: do not DMA-map stack addresses commit 4148df9b0f38bdd362dd91d52076926c11cbe5a9 Author: Paul Mundt Date: Mon Jul 6 00:25:57 2009 +0900 fb: Initialize fb_info mutexes in framebuffer_alloc() This way they'll be properly initialized early enough for users that may touch them before the framebuffer has been registered. Drivers that allocate their fb_info structure some other way (like matrocfb's broken static allocation) need to be fixed up appropriately. Signed-off-by: Paul Mundt Signed-off-by: Linus Torvalds commit 793285fcafce4719a05e0c99fa74b188157fe7fe Author: Oleg Nesterov Date: Sun Jul 5 12:08:26 2009 -0700 cred_guard_mutex: do not return -EINTR to user-space do_execve() and ptrace_attach() return -EINTR if mutex_lock_interruptible(->cred_guard_mutex) fails. This is not right, change the code to return ERESTARTNOINTR. Perhaps we should also change proc_pid_attr_write(). Signed-off-by: Oleg Nesterov Cc: David Howells Acked-by: Roland McGrath Cc: James Morris Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds commit 82e3310ace59794ecf0f531eca94647b2863dfda Author: Tobias Doerffel Date: Sun Jul 5 12:08:23 2009 -0700 linux/sysrq.h needs linux/errno.h In include/linux/sysrq.h the constant EINVAL is being used but is undefined if include/linux/errno.h is not included before. Fix this by adding #include at the beginning. Signed-off-by: Tobias Doerffel Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds commit 600ce1a0faafeed1ce6bcfd421bc040b941cbbc1 Author: InKi Dae Date: Sun Jul 5 12:08:21 2009 -0700 drivers/video/s3c-fb.c: fix clock setting for Samsung SoC Framebuffer Correct the CLKVAL_F field value of VIDEO MAIN CONTROLLER 0 REGITSTER. Frame Rate is 1 / [ { (VSPW+1) + (VBPD+1) + (LIINEVAL + 1) + (VFPD+1) } x {(HSPW+1) + (HBPD +1) + (HFPD+1) + (HOZVAL + 1) } x { ( CLKVAL+1 ) / ( Frequency of Clock source ) } ] and VCLK = Video Clock Source / (CLKVAL +1). therefore CLKVAL_F should be "CLKVAL_F = Frequency of Clock source / pixel clock * refresh". for this, I added refresh value in platform data like below. static struct s3c_fb_pd_win xxx_fb_win0 = { /* this is to ensure we use win0 */ .win_mode = { .refresh = 60, .pixclock = (66+4+2+480)*(15+5+3+800), .left_margin = 66, .right_margin = 2, .upper_margin = 15, .lower_margin = 3, .hsync_len = 4, .vsync_len = 5, .xres = 480, .yres = 800, }, .max_bpp = 32, .default_bpp = 24, }; static struct s3c_fb_platdata xxx_lcd_pdata __initdata = { .win[0] = &xxx_fb_win0, .vidcon0 = VIDCON0_VIDOUT_RGB | VIDCON0_PNRMODE_RGB, .vidcon1 = VIDCON1_INV_HSYNC | VIDCON1_INV_VSYNC | VIDCON1_INV_VCLK | VIDCON1_INV_VDEN, .setup_gpio = s5pc1xx_fb_gpio_setup_24bpp, }; xxx_machine_init() { . . . s3c_fb_set_platdata(&xxx_lcd_pdata); } platform data defined in machine code should be setting using s3c_fb_set_platdata(). Signed-off-by: InKi Dae Cc: Kyungmin Park Cc: Krzysztof Helt Cc: Marek Szyprowski Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds commit 5bfd7560979062ad75c9805c1719cec990b5db29 Author: Kevin Cernekee Date: Sun Jul 5 12:08:19 2009 -0700 Fix virt_to_phys() warnings These warnings were observed on MIPS32 using 2.6.31-rc1 and gcc-4.2.0: mm/page_alloc.c: In function 'alloc_pages_exact': mm/page_alloc.c:1986: warning: passing argument 1 of 'virt_to_phys' makes pointer from integer without a cast drivers/usb/mon/mon_bin.c: In function 'mon_alloc_buff': drivers/usb/mon/mon_bin.c:1264: warning: passing argument 1 of 'virt_to_phys' makes pointer from integer without a cast [akpm@linux-foundation.org: fix kernel/perf_counter.c too] Signed-off-by: Kevin Cernekee Cc: Andi Kleen Cc: Ralf Baechle Cc: Ingo Molnar Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds commit c8236db9cd7aa492dcfcdcca702638e704abed49 Author: Josef Bacik Date: Sun Jul 5 12:08:18 2009 -0700 mm: mark page accessed before we write_end() In testing a backport of the write_begin/write_end AOPs, a 10% re-read regression was noticed when running iozone. This regression was introduced because the old AOPs would always do a mark_page_accessed(page) after the commit_write, but when the new AOPs where introduced, the only place this was kept was in pagecache_write_end(). This patch does the same thing in the generic case as what is done in pagecache_write_end(), which is just to mark the page accessed before we do write_end(). Signed-off-by: Josef Bacik Acked-by: Nick Piggin Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds commit a65e7bfcd74e4c0939f235d2bf9f48ddb3a57991 Author: Hui Zhu Date: Sun Jul 5 12:08:15 2009 -0700 elf: fix multithreaded program core dumping on arm Fix the multithread program core thread message error. This issue affects arches with neither has CORE_DUMP_USE_REGSET nor ELF_CORE_COPY_TASK_REGS, ARM is one of them. The thread message of core file is generated in elf_dump_thread_status. The register values is set by elf_core_copy_task_regs in this function. If an arch doesn't define ELF_CORE_COPY_TASK_REGS, elf_core_copy_task_regs() will do nothing. Then the core file will not have the register message of thread. So add elf_core_copy_regs to set regiser values if ELF_CORE_COPY_TASK_REGS doesn't define. The following is how to reproduce this issue: cat 1.c #include #include #include void td1(void * i) { while (1) { printf ("1\n"); sleep (1); } return; } void td2(void * i) { while (1) { printf ("2\n"); sleep (1); } return; } int main(int argc,char *argv[],char *envp[]) { pthread_t t1,t2; pthread_create(&t1, NULL, (void*)td1, NULL); pthread_create(&t2, NULL, (void*)td2, NULL); sleep (10); assert(0); return (0); } arm-xxx-gcc -g -lpthread 1.c -o 1 copy 1.c and 1 to a arm board. Goto this board. ulimit -c 1800000 ./1 # ./1 1 2 1 ... ... 1 1: 1.c:37: main: Assertion `0' failed. Aborted (core dumped) Then you can get a core file. gdb 1 core.xxx Without the patch: (gdb) info threads 3 process 909 0x00000000 in ?? () 2 process 908 0x00000000 in ?? () * 1 process 907 0x4a6e2238 in raise () from /lib/libc.so.6 You can found that the pc of 909 and 908 is 0x00000000. With the patch: (gdb) info threads 3 process 885 0x4a749974 in nanosleep () from /lib/libc.so.6 2 process 884 0x4a749974 in nanosleep () from /lib/libc.so.6 * 1 process 883 0x4a6e2238 in raise () from /lib/libc.so.6 The pc of 885 and 884 is right. Signed-off-by: Hui Zhu Cc: Amerigo Wang Cc: Al Viro Cc: David Howells Cc: Roland McGrath Cc: Jakub Jelinek Cc: Russell King Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds commit 3beab0b42413e83a7907db7176b54c840fc75a81 Author: Zhang, Yanmin Date: Sun Jul 5 12:08:08 2009 -0700 sys_sync(): fix 16% performance regression in ffsb create_4k test I run many ffsb test cases on JBODs (typically 13/12 disks). Comparing with kernel 2.6.30, 2.6.31-rc1 has about 16% regression with ffsb_create_4k. The sub test case creates files continuously for 10 minitues and every file is 1MB. Bisect located below patch. 5cee5815d1564bbbd505fea86f4550f1efdb5cd0 is first bad commit commit 5cee5815d1564bbbd505fea86f4550f1efdb5cd0 Author: Jan Kara Date: Mon Apr 27 16:43:51 2009 +0200 vfs: Make sys_sync() use fsync_super() (version 4) It is unnecessarily fragile to have two places (fsync_super() and do_sync()) doing data integrity sync of the filesystem. Alter __fsync_super() to accommodate needs of both callers and use it. So after this patch __fsync_super() is the only place where we gather all the calls needed to properly send all data on a filesystem to disk. As a matter of fact, ffsb calls sys_sync in the end to make sure all data is flushed to disks and the flushing is counted into the result. vmstat shows ffsb is blocked when syncing for a long time. With 2.6.30, ffsb is blocked for a short time. I checked the patch and did experiments to recover the original methods. Eventually, the root cause is the patch deletes the calling to wakeup_pdflush when syncing, so only ffsb is blocked on disk I/O. wakeup_pdflush could ask pdflush to write back pages with ffsb at the same time. [akpm@linux-foundation.org: restore comment too] Signed-off-by: Zhang Yanmin Cc: Jan Kara Cc: Al Viro Acked-by: Jens Axboe Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds commit f386c61fe1a1f36f0e434f1b577e6b112698caf7 Author: Peter Oberparleiter Date: Sun Jul 5 12:08:06 2009 -0700 gcov: exclude code operating in userspace from profiling Fix for this issue on x86_64: rostedt@goodmis.org wrote: > On bootup of the latest kernel my init segfaults. Debugging it, > I found that vread_tsc (a vsyscall) increments some strange > kernel memory: > > 0000000000000000 : > 0: 55 push %rbp > 1: 48 ff 05 00 00 00 00 incq 0(%rip) > # 8 > 4: R_X86_64_PC32 .bss+0x3c > 8: 48 89 e5 mov %rsp,%rbp > b: 66 66 90 xchg %ax,%ax > e: 48 ff 05 00 00 00 00 incq 0(%rip) > # 15 > 11: R_X86_64_PC32 .bss+0x44 > 15: 66 66 90 xchg %ax,%ax > 18: 48 ff 05 00 00 00 00 incq 0(%rip) > # 1f > 1b: R_X86_64_PC32 .bss+0x4c > 1f: 0f 31 rdtsc > > > Those "incq" is very bad to happen in vsyscall memory, since > userspace can not modify it. You need to make something prevent > profiling of vsyscall memory (like I do with ftrace). Signed-off-by: Peter Oberparleiter Cc: Ingo Molnar Reported-by: Steven Rostedt Tested-by: Steven Rostedt Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds commit 28df30e61b46a33d1f0bb60757747396886ef687 Author: Jiri Slaby Date: Sun Jul 5 12:08:03 2009 -0700 rtc: ds1374, fix lock imbalance When i2c_smbus_read_byte_data fails in ds1374_work, we forgot to unlock the held lock. Fix that. Signed-off-by: Jiri Slaby Cc: Alessandro Zummo Cc: Scott Wood Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds commit efc0cfa6d16103bd72a7c398f1321816b635370f Author: Florian Fainelli Date: Sun Jul 5 12:08:02 2009 -0700 vlynq: fix typo in Kconfig to enable debugging Fix a typo in the VLYNQ bus driver Kconfig which prevented to turn on VLYNQ bus debugging. Signed-off-by: Florian Fainelli Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds commit 9e2db5c90a25acf7657edb3687b8d5813ccddfee Author: Robert P. J. Day Date: Sun Jul 5 12:08:00 2009 -0700 vlynq: correct typo of missing "CONFIG_" prefix in ifdef Fix a typo in the vlynq bus driver which was missing the CONFIG_ prefix to turn on debugging code. Signed-off-by: Robert P. J. Day Signed-off-by: Florian Fainelli Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds commit cbad1cbb51d92c0b5c3f47ef9dbf125de81fae08 Author: Krzysztof Helt Date: Mon Jul 6 21:11:24 2009 +0200 sisfb: fix regression with uninitalized fb_info->mm_lock mutex Remove redundant call to the sisfb_get_fix() before sis frambuffer is registered. This fixes a problem with uninitialized the fb_info->mm_lock mutex introduced by the commit 537a1bf059f " fbdev: add mutex for fb_mmap locking" Signed-off-by: Krzysztof Helt Tested-by: Wu Zhangjin Signed-off-by: Linus Torvalds commit db6e3f91efd2cf61b9965f722902199cf54adc4f Author: Michal Simek Date: Mon Jul 6 08:21:09 2009 +0200 microblaze: Fix cast warning for init.c Signed-off-by: Michal Simek commit 4ae78338157a1fe2d7a398d86bb9f9043d600d2b Author: Michal Simek Date: Wed Jul 1 15:32:14 2009 +0200 microblaze: Wire up new syscalls Wire up new syscalls rt_tgsigqueueinfo and perf_counter_open. Signed-off-by: Michal Simek commit e513588f75912f022677866244de6b19b98b8d9a Author: Arnd Bergmann Date: Thu Jun 18 19:55:30 2009 +0200 microblaze: use generic syscalls.h The prototypes in syscalls.h all make sense for microblaze, but for some of them, the actual implementation in sys_microblaze.c needs to be adapted. Signed-off-by: Arnd Bergmann Signed-off-by: Michal Simek commit 3183e06863f49a500fc76427db4d60825a26f81b Author: Arnd Bergmann Date: Thu Jun 18 19:55:29 2009 +0200 microblaze: clean up signal handling When legacy signal handling is disabled, the arch/microblaze/kernel/signal.c implementation can be much simpler, as most of it is handled generically from kernel/signal.c. This is also a prerequisite for using the generic asm/unistd.h, which does not provide __NR_sigreturn, because this macro is referenced by the current signal.c implementation. Signed-off-by: Arnd Bergmann Signed-off-by: Michal Simek commit 0a58458341fd571e521be542ff746a4a8995980c Author: Remis Lima Baima Date: Thu Jun 18 19:55:33 2009 +0200 microblaze: convert all simple headers to use asm-generic All the simple microblaze header files were adapted to use their asm-generic implementations. These files are more simple and were quite straightforward to change. fb.h, vga.h and parport.h previously did not exist, using the generic version makes it possible to build more drivers successfully in allyesonfig. Signed-off-by: Remis Lima Baima Signed-off-by: Arnd Bergmann Signed-off-by: Michal Simek commit 14f8738976991d2f8fb6ab6e10e9003562c3d6bb Author: Remis Lima Baima Date: Thu Jun 18 19:55:32 2009 +0200 microblaze: use the generic lib/checksum.c The microblaze checksum code is mostly identical to the asm-generic+lib version, so use that instead. Signed-off-by: Remis Lima Baima Signed-off-by: Michal Simek commit 81d8279ea31a3fc6d4ffacd87119a04c561ca62e Author: Arnd Bergmann Date: Thu Jun 18 19:55:27 2009 +0200 microblaze: fall back on generic header files for the ABI Almost all of the ABI relevant header files now have generic versions, so use those now in order to reduce the amount of architecture specific code. Signed-off-by: Arnd Bergmann Signed-off-by: Michal Simek commit 67fc25ef34a27dc43ce6dfc98a8b249db304d641 Merge: 8e4a718 ec5a36f Author: Pekka Enberg Date: Mon Jul 6 10:51:54 2009 +0300 Merge branch 'slab/urgent' into for-linus commit 0a6843483c256c859cd9542361812a29403f0fb5 Author: Andy Walls Date: Sun Jul 5 16:22:45 2009 -0300 V4L/DVB (12206): get_dvb_firmware: Correct errors in MPC718 firmware extraction logic The extraction routine for the MPC718 "firmware" had 2 bugs in it, where one bug masked the effect of the other. The loop iteration should have set $prevlen = $currlen at the end of the loop, and the if() check should have used && instead of || for deciding if the firmware length is reasonable. Signed-off-by: Andy Walls Signed-off-by: Mauro Carvalho Chehab commit a5ca3a1bc719cb9056acc6f24340a1e2674f21a0 Author: Jiri Slaby Date: Sun Jul 5 15:44:20 2009 -0300 V4L/DVB (12203): radio-si470x: fix lock imbalance There is one path with omitted unlock in si470x_fops_release. Fix that. Cc: Tobias Lorenz Signed-off-by: Jiri Slaby Signed-off-by: Douglas Schilling Landgraf Signed-off-by: Mauro Carvalho Chehab commit 11b79ea75ada39b7f1efdebdad520c93c3ac1f0e Author: Jiri Slaby Date: Sun Jul 5 14:06:41 2009 -0300 V4L/DVB (12202): em28xx, fix lock imbalance There is one omitted unlock in em28xx_usb_probe. Fix that. Signed-off-by: Jiri Slaby Signed-off-by: Douglas Schilling Landgraf Signed-off-by: Mauro Carvalho Chehab commit 791a08fc01aaa293a73c9dce260327fdee288faf Author: Mauro Carvalho Chehab Date: Fri Jul 3 15:36:18 2009 -0300 V4L/DVB (12172): em28xx: Add autodetection code for Silvercrest 1.3 mpix Signed-off-by: Mauro Carvalho Chehab commit 58fc1ce37aba8e6371e1ec8a90d650b1965ee6c8 Author: Mauro Carvalho Chehab Date: Fri Jul 3 02:54:18 2009 -0300 V4L/DVB (12171): em28xx: fix webcam usage with different output formats Discovered the bug that were limiting the output format to just RGB565. Now, it is possible to output image at Bayer format (the original one, as generated by Silvercrest sensor, and two others), and also on YUY. Adds Bayer formats also to the driver. Signed-off-by: Mauro Carvalho Chehab commit ed5f1431ebeeba8cc6739e9cd905a7895b66184c Author: Mauro Carvalho Chehab Date: Thu Jul 2 17:34:04 2009 -0300 V4L/DVB (12169): em28xx-video: fix VIDIOC_G_FMT and VIDIOC_ENUMFMT with webcams Webcams have different constraints than other v4l devices. This patch makes the format ioctls to behave better. It also fixes a bug at open() handler, that were always reseting resolution to the maximum available one. Signed-off-by: Mauro Carvalho Chehab commit f797608cc4f19b44b83ec39c87e14af6fa07826d Author: Devin Heitmueller Date: Mon Jun 22 22:32:32 2009 -0300 V4L/DVB (12156): em28xx: Fix tuning for Terratec Cinergy T XS USB (zl10353 version) Fix the code so that the zl10353 version of the Terratec Cinergy T XS USB starts working again. This includes fixing what must have been a typo in the GPIO definition for the digital side of the board, and setting of the disable_i2c_gate_ctrl property for the zl10353 config, so that the i2c bus doesn't get wedged the first time something tries to close the gate. Also, add a printk() making clear that the mt352 version still isn't supported. This issue is still being actively debugged, but in the meantime at least the dmesg output will show a very clear error... Thanks to Jelle de Jong for providing sample hardware to test with. Thanks to Simon Kenyon for testing various patches and providing SSH access to his environment so I could debug with access to a valid signal source. Cc: Jelle de Jong Cc: Simon Kenyon Signed-off-by: Devin Heitmueller Signed-off-by: Mauro Carvalho Chehab commit 43cb9fe3291bb96390c1d188eb61c2d1581bb61e Author: Mauro Carvalho Chehab Date: Tue Jun 30 08:36:17 2009 -0300 V4L/DVB (12139): em28xx: add other video formats Add suppport for the teste RGB565 format (16 bits/pixel). Currently, webcam support works only at RGB565, at 640x480. Signed-off-by: Mauro Carvalho Chehab commit 02e7804b2135ff941b8846f5820cf48fbfdadd54 Author: Mauro Carvalho Chehab Date: Mon Jun 29 11:35:05 2009 -0300 V4L/DVB (12138): em28xx: add support for Silvercrest Webcam This webcam uses a em2710 chipset, that identifies itself as em2820, plus a mt9v011 sensor, and a DY-301P lens. It needs a few different initializations than a normal em28xx device. Thanks to Hans de Goede and Douglas Landgraf for providing the acces for the webcam during this weekend, I could make a patch for it while returning back from FISL/Fudcom LATAM 2009. Signed-off-by: Mauro Carvalho Chehab commit 6934e6fface7b7d5fe7e578c3f43cbe8f23af2d4 Author: Mauro Carvalho Chehab Date: Sat Jul 4 08:15:11 2009 -0300 V4L/DVB (12174): mt9v011: let's stick with datasheet values where it works The original driver for Silvercrest cameras were using some values that are different from what datasheet says. As result, it was taken very less snapshots per second than expected. A test with the datasheet values showed that they work fine and give a better frame rate. So, let's stick with datasheet values. Signed-off-by: Mauro Carvalho Chehab commit 27fe4a30a211a0561c72b0e519997e91fa91c452 Author: Mauro Carvalho Chehab Date: Sat Jul 4 08:03:48 2009 -0300 V4L/DVB (12173): mt9v011: properly calculate image resolution registers Instead of working with a table of precalculated values, fill them with the proper values. Also, adds format functions that allow changing the resolution, by cropping the image to the center of the sensor. While here, move the sensor version check to the probe routine, to indicate to the caller if the sensor is not supported by this driver. Also, fixes a stupid bug where we're using &buffer[] instead of buffer[]. Signed-off-by: Mauro Carvalho Chehab commit fbe2800c932573e90e38a9c703165839e0c00515 Author: Mauro Carvalho Chehab Date: Mon Jun 29 11:12:05 2009 -0300 V4L/DVB (12137): mt9v011: CodingStyle fixes Signed-off-by: Mauro Carvalho Chehab commit afe09f821ffa733d1cbfbf4e7cc41bd1cb8dcffb Author: Mauro Carvalho Chehab Date: Mon Jun 29 11:03:22 2009 -0300 V4L/DVB (12136): mt9v011: Some fixes at the register initialization table Signed-off-by: Mauro Carvalho Chehab commit 7dfba00d05f3c7db9510f3b54a472981cf1521af Author: Mauro Carvalho Chehab Date: Mon Jun 29 05:41:26 2009 -0300 V4L/DVB (12135): Add a driver for mt9v011 sensor Adds driver for mt9v011 based on its datasheet, available at: http://download.micron.com/pdf/datasheets/imaging/MT9V011.pdf The driver was tested with a webcam that will be added on a next patch. Signed-off-by: Mauro Carvalho Chehab commit b179bc4579f67c6f1df524c48b28cacf0c7a1b91 Author: Michael Krufky Date: Tue Jun 16 17:08:17 2009 -0300 V4L/DVB (12166): cx23885: add FIXME comment above set_frontend override add FIXME comment to indicate that the set_frontend override is a temporary hack. This will be done a better way in the next kernel. Signed-off-by: Michael Krufky Signed-off-by: Mauro Carvalho Chehab commit f35b9e8039c04f0cb9fe3a051ce941b1524a8798 Author: Michael Krufky Date: Fri May 8 22:39:24 2009 -0300 V4L/DVB (12165): cx23885: override set_frontend to allow rf input path switching on the HVR1275 Use separate RF input spigots for Antennae and Cable. Reviewed-by: Steven Toth Signed-off-by: Michael Krufky Signed-off-by: Mauro Carvalho Chehab commit 509dd025a4ac6f32921211557cfd434b81d8d7a7 Author: Mauro Carvalho Chehab Date: Tue Jun 30 16:07:26 2009 -0300 V4L/DVB (12148): move V4L2_PIX_FMT_SGRBG8 to the proper place Instead of defining a new pif format on an internal header, move it to the V4L2 API header. Signed-off-by: Mauro Carvalho Chehab commit ff861fb21e44ab178eddc6a1dd7c5478e3eb2ec5 Author: Andy Walls Date: Fri Jul 3 11:54:40 2009 -0300 V4L/DVB (12182): cx18: Add DVB-T support for Yuan MPC-718 cards with an MT352 or ZL10353 Add DVB-T support for Yuan MPC-718 cards with an MT352 or ZL10353 demodulator. There are apparently some units with a DiBcom demodulator which could be supported by one of the dib7000 modules - but this is not implemented in the cx18 driver (yet). Due to lack of porgramming details for the MT352 and the mt352 module requiring a "demod_init" function, a "firmware" must be obtained and loaded to get DVB-T working for Yuan MPC-718 cards with an MT352. Tested-by: Steve Firth Signed-off-by: Andy Walls Signed-off-by: Mauro Carvalho Chehab commit 9f38a920b232800fd4000ba3d4617b41198e017e Author: Andy Walls Date: Fri Jul 3 11:33:17 2009 -0300 V4L/DVB (12181): get_dvb_firmware: Add Yuan MPC718 MT352 DVB-T "firmware" extraction Add routine to support extracting the MT352 DVB-T demodulator initialization sequence for Yuan MPC718 cards for use by the cx18 driver. This routine uses a hueristic for extracting a good sequence. It should work on all different versions of the "yuanrap.sys" file, given the way the MT352 tuning sequences are stored in all versions of that file I have seen so far. However, the current patch simply looks for one specific archive URL. Signed-off-by: Andy Walls Signed-off-by: Mauro Carvalho Chehab commit 637fb3d70b1e53a127cba5ffae5aa3b2b9f57df6 Author: Andy Walls Date: Fri Jul 3 11:30:27 2009 -0300 V4L/DVB (12180): cx18: Update Yuan MPC-718 card entry with better information and guesses Signed-off-by: Andy Walls Signed-off-by: Mauro Carvalho Chehab commit 6e707b4c6d1fcf0fca6a4525e0ff66d35e21a517 Author: Andy Walls Date: Thu Jun 11 07:57:50 2009 -0300 V4L/DVB (12167): tuner-xc2028: Fix 7 MHz DVB-T The following patch should fix 7 MHz DVB-T with the XC3028 using the DTV7 firmware from the xc3028-v27.fw firmware image. Tested-by: Terry Wu [mchehab@redhat.com: thanks to John Ferlito, we have a real test in Australia] Signed-off-by: Andy Walls Signed-off-by: Mauro Carvalho Chehab commit c06950ae4dccf59566fec7ae269aaeb24685c16c Author: Guennadi Liakhovetski Date: Wed Jun 24 10:31:25 2009 -0300 V4L/DVB (12160): soc-camera: fix missing clean up on error path If soc_camera_init_user_formats() fails in soc_camera_probe(), we have to call client's .remove() method to unregister the video device. Reported-by: Kuninori Morimoto Signed-off-by: Guennadi Liakhovetski Signed-off-by: Mauro Carvalho Chehab commit 14df2ccead8f6da5a9c8f904fc629aaf849b7e55 Author: Stefan Herbrechtsmeier Date: Wed Jun 24 10:31:25 2009 -0300 V4L/DVB (12159): soc_camera: Fix debug output of supported formats count The supported formats count must be set to 0 after debug output right before the second pass. Signed-off-by: Stefan Herbrechtsmeier Signed-off-by: Guennadi Liakhovetski Signed-off-by: Mauro Carvalho Chehab commit 3175da83a0a85fa03c9c34061906a9ca0725af0d Author: Hans Verkuil Date: Fri Jun 26 15:35:04 2009 -0300 V4L/DVB (12153): ttpci: config TTPCI_EEPROM depends on I2C If I2C is not enabled, then we shouldn't build ttpci_eeprom.c. Signed-off-by: Hans Verkuil Signed-off-by: Mauro Carvalho Chehab commit c285addb3991c9b73b8d1a2b652ce0f1f67e90e1 Author: Mauro Carvalho Chehab Date: Thu Jun 25 16:28:23 2009 -0300 V4L/DVB (12134): vivi: bug: don't assume that S_STD will be called before streaming precalculate_bars() improved vivi performance. However, it assumed that always before streaming, the driver would call VIDIOC_S_STD. This is not an API requirement, and the testing apps don't do that. Due to that, a regression were caused by the patch that added it. This patch moves the precalculate_bars to the proper place of the code, calling it at buffer_prepare() callback. Signed-off-by: Mauro Carvalho Chehab commit 3dfc813d94bba2046c6aed216e0fd69ac93a8e03 Author: David Woodhouse Date: Sat Jul 4 19:11:08 2009 +0100 intel-iommu: Don't use identity mapping for PCI devices behind bridges Our current strategy for pass-through mode is to put all devices into the 1:1 domain at startup (which is before we know what their dma_mask will be), and only _later_ take them out of that domain, if it turns out that they really can't address all of memory. However, when there are a bunch of PCI devices behind a bridge, they all end up with the same source-id on their DMA transactions, and hence in the same IOMMU domain. This means that we _can't_ easily move them from the 1:1 domain into their own domain at runtime, because there might be DMA in-flight from their siblings. So we have to adjust our pass-through strategy: For PCI devices not on the root bus, and for the bridges which will take responsibility for their transactions, we have to start up _out_ of the 1:1 domain, just in case. This fixes the BUG() we see when we have 32-bit-capable devices behind a PCI-PCI bridge, and use the software identity mapping. It does mean that we might end up using 'normal' mapping mode for some devices which could actually live with the faster 1:1 mapping -- but this is only for PCI devices behind bridges, which presumably aren't the devices for which people are most concerned about performance. Signed-off-by: David Woodhouse commit 6941af2810c6fc970b88f7c0d52ba4e286acbee5 Author: David Woodhouse Date: Sat Jul 4 18:24:27 2009 +0100 intel-iommu: Use iommu_should_identity_map() at startup time too. At boot time, the dma_mask won't have been set on any devices, so we assume that all devices will be 64-bit capable (and thus get a 1:1 map). Signed-off-by: David Woodhouse commit 736768325efcbee7b0861d62670d01a54c2d158b Author: David Woodhouse Date: Sat Jul 4 14:08:36 2009 +0100 intel-iommu: No mapping for non-PCI devices This should fix kernel.org bug #11821, where the dcdbas driver makes up a platform device and then uses dma_alloc_coherent() on it, in an attempt to get memory < 4GiB. Signed-off-by: David Woodhouse commit 62edf5dc4a524e4cb525e6020b955a1ad593d9ba Author: David Woodhouse Date: Sat Jul 4 10:59:46 2009 +0100 intel-iommu: Restore DMAR_BROKEN_GFX_WA option for broken graphics drivers We need to give people a little more time to fix the broken drivers. Re-introduce this, but tied in properly with the 'iommu=pt' support this time. Change the config option name and make it default to 'no' too. Signed-off-by: David Woodhouse commit 40e4aa34324bff3938a900014254f88943d05e15 Author: David Woodhouse Date: Sat Jul 4 10:55:41 2009 +0100 intel-iommu: Add iommu_should_identity_map() function We do this twice, and it's about to get more complicated. This makes the code slightly clearer about what it's doing, too. Signed-off-by: David Woodhouse commit 1b7bc0a1618b4de1e6f55c6d95b790f4ab6fcd9e Author: David Woodhouse Date: Sat Jul 4 10:49:46 2009 +0100 intel-iommu: Fix reattaching of devices to identity mapping domain When we reattach a device to the si_domain (because it's been removed from a VM), we weren't calling domain_context_mapping() to actually tell the hardware about that. We should really put the call to domain_context_mapping() into domain_add_dev_info() -- we never call the latter without also doing the former, and we can keep the error paths simple that way. But that's a cleanup which can wait for 2.6.32 now. Signed-off-by: David Woodhouse commit 1e4c64c46d413de84cc0b786bd6a9b555ba7d111 Author: David Woodhouse Date: Sat Jul 4 10:40:38 2009 +0100 intel-iommu: Don't set identity mapping for bypassed graphics devices We should check iommu_dummy() _first_, because that means it's attached to an iommu that we've just disabled completely. At the moment, we might try to put the device into the identity mapping domain. Signed-off-by: David Woodhouse commit 5a5e02a614e59db7536cd11029e6674adc41b191 Author: David Woodhouse Date: Sat Jul 4 09:35:44 2009 +0100 intel-iommu: Fix dma vs. mm page confusion with aligned_nrpages() The aligned_nrpages() function rounds up to the next VM page, but returns its result as a number of DMA pages. Purely theoretical except on IA64, which doesn't boot with VT-d right now anyway. Signed-off-by: David Woodhouse commit ebbb16bffa646f853899ef3fdc0ac7abab888703 Author: Stefan Richter Date: Tue Jun 30 20:28:31 2009 +0200 ieee1394: sbp2: add support for disks >2 TB (and 16 bytes long CDBs) Increase the command ORB data structure to transport up to 16 bytes long CDBs (instead of 12 bytes), and tell the SCSI mid layer about it. This is notably necessary for READ CAPACITY(16) and friends, i.e. support of large disks. Signed-off-by: Stefan Richter commit af2719415a5ceae06f2a6d33e78b555e64697fc8 Author: Stefan Richter Date: Tue Jun 30 20:27:59 2009 +0200 firewire: sbp2: add support for disks >2 TB (and 16 bytes long CDBs) Increase the command ORB data structure to transport up to 16 bytes long CDBs (instead of 12 bytes), and tell the SCSI mid layer about it. This is notably necessary for READ CAPACITY(16) and friends, i.e. support of large disks. Signed-off-by: Stefan Richter commit 46b952a3c3a94afa339bd4961a4f3d1482436599 Author: Matthew Wilcox Date: Wed Jul 1 14:24:30 2009 -0700 PCI: Fix IRQ swizzling for ARI-enabled devices For many purposes, including interrupt-swizzling, devices with ARI enabled behave as if they have one device (number 0) and 256 functions. This probably hasn't bitten us in practice because all ARI devices I've seen are also IOV devices, and IOV devices are required to use MSI. This isn't guaranteed, and there are legitimate reasons to use ARI without IOV, and hence potentially use pin-based interrupts. Signed-off-by: Matthew Wilcox Signed-off-by: Jesse Barnes commit 944c54e7fc5ccf961bef2b5449958436b85de459 Author: Ingo Molnar Date: Wed Jul 1 00:10:16 2009 +0200 ia64/PCI: adjust section annotation for pcibios_setup() Should be __init. Acked-by: Tony Luck Signed-off-by: Jesse Barnes commit 626fdfec1588ac1341a37805809d03a719d977e0 Author: Yinghai Lu Date: Wed Jun 24 20:00:12 2009 -0700 x86/PCI: get root CRS before scanning children This allows us to remove adjust_transparent_bridge_resources and give x86_pci_root_bus_res_quirks a chance when _CRS is not used or not there. Acked-by: Gary Hade Tested-by: Gary Hade Signed-off-by: Yinghai Lu Signed-off-by: Jesse Barnes commit 2cdb3f1d834aab27a927be7555fbf4f9e43e9261 Author: Yinghai Lu Date: Wed Jun 24 19:01:19 2009 -0700 x86/PCI: fix boundary checking when using root CRS Don't touch info->res_num if we are out of space. Acked-by: Gary Hade Tested-by: Gary Hade Signed-off-by: Yinghai Lu Signed-off-by: Jesse Barnes commit 12abb8ba8444f7c9b355bbdd44a6d0839f7a41b6 Author: Hidetoshi Seto Date: Wed Jun 24 12:08:09 2009 +0900 PCI MSI: Fix restoration of MSI/MSI-X mask states in suspend/resume There are 2 problems on mask states in suspend/resume. [1]: It is better to restore the mask states of MSI/MSI-X to initial states (MSI is unmasked, MSI-X is masked) when we release the device. The pci_msi_shutdown() does the restoration of mask states for MSI, while the msi_free_irqs() does it for MSI-X. In other words, in the "disable" path both of MSI and MSI-X are handled, but in the "shutdown" path only MSI is handled. MSI: pci_disable_msi() => pci_msi_shutdown() [ mask states for MSI restored ] => msi_set_enable(dev, pos, 0); => msi_free_irqs() MSI-X: pci_disable_msix() => pci_msix_shutdown() => msix_set_enable(dev, 0); => msix_free_all_irqs => msi_free_irqs() [ mask states for MSI-X restored ] This patch moves the masking for MSI-X from msi_free_irqs() to pci_msix_shutdown(). This change has some positive side effects: - It prevents OS from touching mask states before reading preserved bits in the register, which can be happen if msi_free_irqs() is called from error path in msix_capability_init(). - It also prevents touching the register after turning off MSI-X in "disable" path, which can be a problem on some devices. [2]: We have cache of the mask state in msi_desc, which is automatically updated when msi/msix_mask_irq() is called. This cached states are used for the resume. But since what need to be restored in the resume is the states before the shutdown on the suspend, calling msi/msix_mask_irq() from pci_msi/msix_shutdown() is not appropriate. This patch introduces __msi/msix_mask_irq() that do mask as same as msi/msix_mask_irq() but does not update cached state, for use in pci_msi/msix_shutdown(). [updated: get rid of msi/msix_mask_irq_nocache() (proposed by Matthew Wilcox)] Reviewed-by: Matthew Wilcox Signed-off-by: Hidetoshi Seto Signed-off-by: Jesse Barnes commit 7ba1930db02fc3118165338ef4e562869f575583 Author: Hidetoshi Seto Date: Tue Jun 23 17:39:27 2009 +0900 PCI MSI: Unmask MSI if setup failed The initial state of mask register of MSI is unmasked. We set it masked before calling arch_setup_msi_irqs(). If arch_setup_msi_irq() fails, it is better to restore the state of the mask register. Reviewed-by: Matthew Wilcox Signed-off-by: Hidetoshi Seto Signed-off-by: Jesse Barnes commit 2c21fd4b333e4c780a46edcd6d1e85bfc6cdf371 Author: Hidetoshi Seto Date: Tue Jun 23 17:40:04 2009 +0900 PCI MSI: shorten PCI_MSIX_ENTRY_* symbol names These names are too long! Drop _OFFSET to save some bytes/lines. Reviewed-by: Matthew Wilcox Signed-off-by: Hidetoshi Seto Signed-off-by: Jesse Barnes commit 2fc90f6133a87da8177636866557d4cc5f56e661 Author: Alexey Zaytsev Date: Wed Jun 24 16:22:30 2009 +0400 PCI: make pci_name() take const argument Since this function should never modify it (saves warnings when called with const args too). Signed-off-by: Alexey Zaytsev Signed-off-by: Jesse Barnes commit 7a661c6f1082693a7e9627e9ad2d1546a9337fdc Author: Alan Cox Date: Wed Jun 24 16:02:27 2009 +0100 PCI: More PATA quirks for not entering D3 The ALi loses some state if it goes into D3. Unfortunately even with the chipset documents I can't figure out how to restore some bits of it. The VIA one saves/restores apparently fine but the ACPI _GTM methods break on some platforms if we do this and this causes cable misdetections. These are both effectively regressions as historically nothing matched the devices and then decided not to bind to them. Nowdays something is binding to all sorts of devices and a result they get dumped into D3. Signed-off-by: Alan Cox Acked-by: Jeff Garzik Signed-off-by: Jesse Barnes commit 503998ca4a295f7da233689850ba4b9d13cf41e7 Author: Randy Dunlap Date: Wed Jun 24 09:18:14 2009 -0700 PCI: fix kernel-doc warnings Add documentation for missing parameters in PCI hotplug code. Signed-off-by: Randy Dunlap Signed-off-by: Jesse Barnes commit 654b75e044119bf8e7d773bce41ea039281bbfbe Author: Yu Zhao Date: Fri Jun 26 14:04:46 2009 +0800 PCI: check if bus has a proper bridge device before triggering SBR For devices attached to the root bus, we can't trigger Secondary Bus Reset because there is no bridge device associated with the bus. So need to check bus->self again NULL first before using it. Reviewed-by: Kenji Kaneshige Signed-off-by: Yu Zhao Signed-off-by: Jesse Barnes commit 977badb2f385db14d8ba3fcf02fbd7ead8e63d0f Author: Fernando Luis Vázquez Cao Date: Fri Jun 26 11:27:41 2009 +0900 PCI: remove pci_dac_dma_... APIs on mn10300 It seems that mn10300 made it upstream after Jan Beulich's pci_dac_dma_* cleanup work and still defines pci_dac_dma_supported(). This API is not required by the PCI subsystem anymore, so remove it. Acked-by: David Howells Signed-off-by: Fernando Luis Vazquez Cao Signed-off-by: Jesse Barnes commit 50e5628a4ac465a52f0d4ca6567343be029731a0 Author: Joe Perches Date: Sun Jun 28 09:26:40 2009 -0700 PCI ECRC: Remove unnecessary semicolons Acked-by: Andrew Patterson Signed-off-by: Joe Perches Signed-off-by: Jesse Barnes commit 0d07348931daef854aca8c834a89f1a99ba4ff2b Author: Hidetoshi Seto Date: Wed Jun 24 12:08:27 2009 +0900 PCI MSI: Return if alloc_msi_entry for MSI-X failed In current code it continues setup even if alloc_msi_entry() for MSI-X is failed due to lack of memory. It means arch_setup_msi_irqs() might be called with msi_desc entries less than its argument nvec. At least x86's arch_setup_msi_irqs() uses list_for_each_entry() for dev->msi_list that suspected to have entries same numbers as nvec, and it doesn't check the number of allocated vectors and passed arg nvec. Therefore it will result in success of pci_enable_msix(), with less vectors allocated than requested. This patch fixes the error route to return -ENOMEM, instead of continuing the setup (proposed by Matthew Wilcox). Note that there is no iounmap in msi_free_irqs() if no msi_disc is allocated. Reviewed-by: Matthew Wilcox Signed-off-by: Hidetoshi Seto Signed-off-by: Jesse Barnes commit ec5a36f94e7ca4b1f28ae4dd135cd415a704e772 Author: Pekka Enberg Date: Mon Jun 29 09:57:10 2009 +0300 SLAB: Fix lockdep annotations Commit 8429db5... ("slab: setup cpu caches later on when interrupts are enabled") broke mm/slab.c lockdep annotations: [ 11.554715] ============================================= [ 11.555249] [ INFO: possible recursive locking detected ] [ 11.555560] 2.6.31-rc1 #896 [ 11.555861] --------------------------------------------- [ 11.556127] udevd/1899 is trying to acquire lock: [ 11.556436] (&nc->lock){-.-...}, at: [] kmem_cache_free+0xcd/0x25b [ 11.557101] [ 11.557102] but task is already holding lock: [ 11.557706] (&nc->lock){-.-...}, at: [] kfree+0x137/0x292 [ 11.558109] [ 11.558109] other info that might help us debug this: [ 11.558720] 2 locks held by udevd/1899: [ 11.558983] #0: (&nc->lock){-.-...}, at: [] kfree+0x137/0x292 [ 11.559734] #1: (&parent->list_lock){-.-...}, at: [] __drain_alien_cache+0x3b/0xbd [ 11.560442] [ 11.560443] stack backtrace: [ 11.561009] Pid: 1899, comm: udevd Not tainted 2.6.31-rc1 #896 [ 11.561276] Call Trace: [ 11.561632] [] __lock_acquire+0x15ec/0x168f [ 11.561901] [] ? __lock_acquire+0x1676/0x168f [ 11.562171] [] ? trace_hardirqs_on_caller+0x113/0x13e [ 11.562490] [] ? trace_hardirqs_on_thunk+0x3a/0x3f [ 11.562807] [] lock_acquire+0xc1/0xe5 [ 11.563073] [] ? kmem_cache_free+0xcd/0x25b [ 11.563385] [] _spin_lock+0x31/0x66 [ 11.563696] [] ? kmem_cache_free+0xcd/0x25b [ 11.563964] [] kmem_cache_free+0xcd/0x25b [ 11.564235] [] ? __free_pages+0x1b/0x24 [ 11.564551] [] slab_destroy+0x57/0x5c [ 11.564860] [] free_block+0xd8/0x123 [ 11.565126] [] __drain_alien_cache+0xa2/0xbd [ 11.565441] [] kfree+0x14c/0x292 [ 11.565752] [] skb_release_data+0xc6/0xcb [ 11.566020] [] __kfree_skb+0x19/0x86 [ 11.566286] [] consume_skb+0x2b/0x2d [ 11.566631] [] skb_free_datagram+0x14/0x3a [ 11.566901] [] netlink_recvmsg+0x164/0x258 [ 11.567170] [] sock_recvmsg+0xe5/0xfe [ 11.567486] [] ? might_fault+0xaf/0xb1 [ 11.567802] [] ? autoremove_wake_function+0x0/0x38 [ 11.568073] [] ? core_sys_select+0x3d/0x2b4 [ 11.568378] [] ? __lock_acquire+0x1676/0x168f [ 11.568693] [] ? sockfd_lookup_light+0x1b/0x54 [ 11.568961] [] sys_recvfrom+0xa3/0xf8 [ 11.569228] [] ? trace_hardirqs_on+0xd/0xf [ 11.569546] [] system_call_fastpath+0x16/0x1b# Fix that up. Closes-bug: http://bugzilla.kernel.org/show_bug.cgi?id=13654 Tested-by: Venkatesh Pallipadi Signed-off-by: Pekka Enberg commit 7ed9f7e5db58c6e8c2b4b738a75d5dcd8e17aad5 Author: Paul E. McKenney Date: Thu Jun 25 12:31:37 2009 -0700 fix RCU-callback-after-kmem_cache_destroy problem in sl[aou]b Jesper noted that kmem_cache_destroy() invokes synchronize_rcu() rather than rcu_barrier() in the SLAB_DESTROY_BY_RCU case, which could result in RCU callbacks accessing a kmem_cache after it had been destroyed. Cc: Acked-by: Matt Mackall Reported-by: Jesper Dangaard Brouer Signed-off-by: Paul E. McKenney Signed-off-by: Pekka Enberg commit 6fdc03709433ccc2005f0f593ae9d9dd04f7b485 Author: Stefan Richter Date: Sat Jun 20 13:23:59 2009 +0200 firewire: core: do not DMA-map stack addresses The DMA mapping API cannot map on-stack addresses, as explained in Documentation/DMA-mapping.txt. Convert the two cases of on-stack packet payload buffers in firewire-core (payload of lock requests in the bus manager work and in iso resource management) to slab-allocated memory. There are a number on-stack buffers for quadlet write or quadlet read requests in firewire-core and firewire-sbp2. These are harmless; they are copied to/ from card driver internal DMA buffers since quadlet payloads are inlined with packet headers. Signed-off-by: Stefan Richter