summaryrefslogtreecommitdiffstats
AgeCommit message (Collapse)Author
2012-10-11make sure that we always have a return path from kernel_execve()Al Viro
The only place where kernel_execve() is called without a way to return to the caller of kernel_thread() callback is kernel_post(). Reorganize kernel_init()/kernel_post() - instead of the former calling the latter in the end (and getting freed by it), have the latter *begin* with calling the former (and turn the latter into kernel_thread() callback, of course). Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
2012-10-11ppc: eeh_event should just use kthread_run()Al Viro
Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
2012-10-11don't bother with kernel_thread/kernel_execve for launching linuxrcAl Viro
exec_usermodehelper_fns() will do just fine... Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
2012-10-11alpha: get rid of switch_stack argument of do_work_pending()Al Viro
... and now the asm glue side of that. Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
2012-10-11alpha: don't bother passing switch_stack separately from regsAl Viro
It's needed only in setup_sigcontext() and it's always reg - <constant>; no point passing it all way down through the call chain. This is just the signal.c side of that stuff; next will come the asm glue one... Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
2012-10-11alpha: take SIGPENDING/NOTIFY_RESUME loop into signal.cAl Viro
Turn the slow side of work_pending into C function, including all the looping. What we get out of that: * we do _not_ call get_signal_to_deliver() with IRQs disabled anymore * no need to save/restore volatiles on each pass if there turns to be more than one (unlikely, but still) * all double-restart prevention is in C now. * glue gets simpler. Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
2012-10-11alpha: simplify TIF_NEED_RESCHED handlingAl Viro
In case we have both NEED_RESCHED and SIGPENDING/NOTIFY_RESUME, handle the latter first. We'll get to original priorities in the next commit, but now that allows to simplify the treatment of NEED_RESCHED-only case nicely. Namely, now there no need to preserve the data for restarts across the call of schedule() in $work_resched; we can get there only if we had either returned from syscall without SIGPENDING (in which case we should've had no restart-worthy return value and want no restarts) or already got through do_notify_resume() call (in which case we want no restarts anymore). So we can just slap 0 into $19 instead of preserving it (and $20). Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
2012-10-12Merge branch 'misc' of ↵Linus Torvalds
git://git.kernel.org/pub/scm/linux/kernel/git/mmarek/kbuild Pull kbuild misc changes from Michal Marek: "In the non-critical part of kbuild, I have - Some make coccicheck improvements and two new tests - Support for a cleaner html output in scripts/kernel-doc, named html5 (no, it does not play videos, yet) BTW, Randy wants to route further kernel-doc patches through the kbuild tree." * 'misc' of git://git.kernel.org/pub/scm/linux/kernel/git/mmarek/kbuild: Update SmPL/Coccinelle section of MAINTAINERS coccicheck: Add the rep+ctxt mode scripts/coccinelle/tests/odd_ptr_err.cocci: semantic patch for IS_ERR/PTR_ERR inconsistency scripts/tags.sh: Add magic for pci access functions scripts/coccinelle: ptr_ret: Add ternary operator version scripts/kernel-doc: drop maintainer scripts/kernel-doc: added support for html5
2012-10-12Merge branch 'kconfig' of ↵Linus Torvalds
git://git.kernel.org/pub/scm/linux/kernel/git/mmarek/kbuild Pull kconfig changes from Michal Marek: "kconfig in v3.7 is going to - initialize ncurses only once in menuconfig - be able to jump to a search result in menuconfig - change the misnomer oldnoconfig to a more meaningful name olddefconfig, keeping the old name as alias" * 'kconfig' of git://git.kernel.org/pub/scm/linux/kernel/git/mmarek/kbuild: kconfig: replace 'oldnoconfig' with 'olddefconfig', and keep the old name as an alias menuconfig: Assign jump keys per-page instead of globally menuconfig: Do not open code textbox scroll up/down menuconfig: Add jump keys to search results menuconfig: Extend dialog_textbox so that it can return to a scrolled position menuconfig: Extend dialog_textbox so that it can exit on arbitrary keypresses menuconfig: Remove superfluous conditionnal kconfig: document oldnoconfig to what it really does in conf.c kconfig/mconf.c: revision of curses initialization.
2012-10-12Merge branch 'kbuild' of ↵Linus Torvalds
git://git.kernel.org/pub/scm/linux/kernel/git/mmarek/kbuild Pull kbuild changes from Michal Marek: "The main part of kbuild for v3.7 contains: - Fix for scripts/Makefile.modpost to not choke on a '.ko' substring in the build directory path - Two warning fixes (modpost and main Makefile) - __compiletime_error works also with gcc 4.3 - make tar{gz,bz2,xz}-pkg uses default compression settings instead of saving as many bytes as possible (this should actually be in the misc branch, I don't know why I applied it here)." * 'kbuild' of git://git.kernel.org/pub/scm/linux/kernel/git/mmarek/kbuild: compiler-gcc4.h: correct verion check for __compiletime_error modpost: Permit .GCC.command.line sections Kbuild: use normal compression settings for tar*-pkg scripts/Makefile.modpost: error in finding modules from .mod files. kbuild: Remove useless warning while appending KCFLAGS
2012-10-12Merge tag 'fbdev-updates-for-3.7' of git://github.com/schandinat/linux-2.6Linus Torvalds
Pull fbdev updates from Florian Tobias Schandinat: "This includes: - large updates for OMAP - basic OMAP5 DSS support for DPI and DSI outputs - large cleanups and restructuring - some update to Exynos and da8xx-fb - removal of the pnx4008 driver (arch removed) - various other small patches" Fix up some trivial conflicts (mostly just include line changes, but also some due to the renaming of the deferred work functions by Tejun). * tag 'fbdev-updates-for-3.7' of git://github.com/schandinat/linux-2.6: (193 commits) gbefb: fix compile error video: mark nuc900fb_map_video_memory as __devinit video/mx3fb: set .owner to prevent module unloading while being used video: exynos_dp: use clk_prepare_enable and clk_disable_unprepare drivers/video/exynos/exynos_mipi_dsi.c: fix error return code drivers/video/savage/savagefb_driver.c: fix error return code video: s3c-fb: use clk_prepare_enable and clk_disable_unprepare da8xx-fb: save and restore LCDC context across suspend/resume cycle da8xx-fb: add pm_runtime support video/udlfb: fix line counting in fb_write OMAPDSS: add missing include for string.h OMAPDSS: DISPC: Configure color conversion coefficients for writeback OMAPDSS: DISPC: Add manager like functions for writeback OMAPDSS: DISPC: Configure writeback FIFOs OMAPDSS: DISPC: Configure writeback specific parameters in dispc_wb_setup() OMAPDSS: DISPC: Configure overlay-like parameters in dispc_wb_setup OMAPDSS: DISPC: Add function to set channel in for writeback OMAPDSS: DISPC: Don't set chroma resampling bit for writeback OMAPDSS: DISPC: Downscale chroma if plane is writeback OMAPDSS: DISPC: Configure input and output sizes for writeback ...
2012-10-12Merge tag 'for-linus-merge-3.7' of ↵Linus Torvalds
git://git.kernel.org/pub/scm/linux/kernel/git/ericvh/v9fs Pull v9fs update from Eric Van Hensbergen. * tag 'for-linus-merge-3.7' of git://git.kernel.org/pub/scm/linux/kernel/git/ericvh/v9fs: 9P: Fix race between p9_write_work() and p9_fd_request() 9P: Fix race in p9_write_work() 9P: fix test at the end of p9_write_work() 9P: Fix race in p9_read_work() 9p: don't use __getname/__putname for uname/aname net/9p: Check errno validity fs/9p: avoid debug OOPS when reading a long symlink
2012-10-11vfs: bogus warnings in fs/namei.cArnd Bergmann
The follow_link() function always initializes its *p argument, or returns an error, but when building with 'gcc -s', the compiler gets confused by the __always_inline attribute to the function and can no longer detect where the cookie was initialized. The solution is to always initialize the pointer from follow_link, even in the error path. When building with -O2, this has zero impact on generated code and adds a single instruction in the error path for a -Os build on ARM. Without this patch, building with gcc-4.6 through gcc-4.8 and CONFIG_CC_OPTIMIZE_FOR_SIZE results in: fs/namei.c: In function 'link_path_walk': fs/namei.c:649:24: warning: 'cookie' may be used uninitialized in this function [-Wuninitialized] fs/namei.c:1544:9: note: 'cookie' was declared here fs/namei.c: In function 'path_lookupat': fs/namei.c:649:24: warning: 'cookie' may be used uninitialized in this function [-Wuninitialized] fs/namei.c:1934:10: note: 'cookie' was declared here fs/namei.c: In function 'path_openat': fs/namei.c:649:24: warning: 'cookie' may be used uninitialized in this function [-Wuninitialized] fs/namei.c:2899:9: note: 'cookie' was declared here Signed-off-by: Arnd Bergmann <arnd@arndb.de> Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
2012-10-11consitify do_mount() argumentsAl Viro
Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
2012-10-11UAPI: (Scripted) Disintegrate include/linux/mmcDavid Howells
Signed-off-by: David Howells <dhowells@redhat.com> Acked-by: Arnd Bergmann <arnd@arndb.de> Acked-by: Thomas Gleixner <tglx@linutronix.de> Acked-by: Michael Kerrisk <mtk.manpages@gmail.com> Acked-by: Paul E. McKenney <paulmck@linux.vnet.ibm.com> Acked-by: Dave Jones <davej@redhat.com> Signed-off-by: Chris Ball <cjb@laptop.org>
2012-10-11mmc: mxs-mmc: Fix merge issue causing build errorMarek Vasut
The following error appeared due to a merge problem; the patches: fc108d24 "mmc: mxs-mmc: fix deadlock caused by recursion loop" 829c1bf4 "mmc: spi: Pull out parts shared between MMC and SPI" came in through separate branches and cause this build error when combined. drivers/mmc/host/mxs-mmc.c: In function 'mxs_mmc_enable_sdio_irq': drivers/mmc/host/mxs-mmc.c:527:3: error: 'struct mxs_mmc_host' has no member named 'base' drivers/mmc/host/mxs-mmc.c:527:3: error: 'struct mxs_mmc_host' has no member named 'devid' make[3]: *** [drivers/mmc/host/mxs-mmc.o] Error 1 This patch corrects the issue. Signed-off-by: Marek Vasut <marex@denx.de> Cc: Fabio Estevam <fabio.estevam@freescale.com> Cc: Shawn Guo <shawn.guo@linaro.org> Signed-off-by: Chris Ball <cjb@laptop.org>
2012-10-11compiler-gcc4.h: correct verion check for __compiletime_errorDaniel Santos
__attribute__((error(msg))) was introduced in gcc 4.3 (not 4.4) and as I was unable to find any gcc bugs pertaining to it, I'm presuming that it has functioned as advertised since 4.3.0. Signed-off-by: Daniel Santos <daniel.santos@pobox.com> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Tested-by: David Rientjes <rientjes@google.com> Signed-off-by: Michal Marek <mmarek@suse.cz>
2012-10-11modpost: Permit .GCC.command.line sectionsJonathan Kliegman
Allow .GCC.command.line sections in modules to prevent modpost warnings: WARNING: sound/usb/snd-usbmidi-lib.o (.GCC.command.line): unexpected non-allocatable section. Did you forget to use "ax"/"aw" in a .S file? Note that for example <linux/init.h> contains section definitions for use in .S files. Signed-off-by: Jonathan Kliegman <kliegs@chromium.org> Signed-off-by: Michal Marek <mmarek@suse.cz>
2012-10-11usbnet: Support devices reporting idlenessOliver Neukum
Some device types support a form of power management in which the device suggests to the host that the device may be suspended now. Support for that is best located in usbnet. Signed-off-by: Oliver Neukum <oneukum@suse.de> Signed-off-by: David S. Miller <davem@davemloft.net>
2012-10-11Add CDC-ACM support for the CX93010-2x UCMxx USB ModemJean-Christian de Rivaz
This USB V.92/V.32bis Controllered Modem have the USB vendor ID 0x0572 and device ID 0x1340. It need the NO_UNION_NORMAL quirk to be recognized. Reference: http://www.conexant.com/servlets/DownloadServlet/DSH-201723-005.pdf?docid=1725&revid=5 See idVendor and idProduct in table 6-1. Device Descriptors Signed-off-by: Jean-Christian de Rivaz <jc@eclis.ch> Signed-off-by: David S. Miller <davem@davemloft.net>
2012-10-11net/ethernet/jme: disable ASPMKevin Baradon
Based on patch from Matthew Garrett <mjg@redhat.com> (https://lkml.org/lkml/2011/11/11/168). http://driveragent.com/archive/30421/7-0-14 indicates that ASPM is disabled on the 250 and 260. Duplicate for sanity. Fixes random RX engine hangs I experienced with JMC250 on Clevo W270HU. Signed-off-by: Kevin Baradon <kevin.baradon@gmail.com> Cc: Guo-Fu Tseng <cooldavid@cooldavid.org> Cc: Matthew Garrett <mjg@redhat.com> Cc: netdev@vger.kernel.org Signed-off-by: David S. Miller <davem@davemloft.net>
2012-10-11Merge tag 'disintegrate-wimax-20121009' of ↵David S. Miller
git://git.infradead.org/users/dhowells/linux-headers UAPI Disintegration 2012-10-09 Signed-off-by: David S. Miller <davem@davemloft.net>
2012-10-11tcp: sysctl interface leaks 16 bytes of kernel memoryAlan Cox
If the rc_dereference of tcp_fastopen_ctx ever fails then we copy 16 bytes of kernel stack into the proc result. Signed-off-by: Alan Cox <alan@linux.intel.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2012-10-11kaweth: print correct debug ptrAlan Cox
We nowdays copy the buffer and free fw->data, so make the debug printk use the right thing. Signed-off-by: Alan Cox <alan@linux.intel.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2012-10-11Merge branch 'staging/for_v3.7' into v4l_for_linusMauro Carvalho Chehab
Applied on the top of changeset 782cd9e, as some of those patches depend on some fixes that went via -arm tree. * staging/for_v3.7: (109 commits) [media] m5mols: Add missing #include <linux/sizes.h> [media] stk1160: Add support for S-Video input Revert "[media] omap3isp: Replace cpu_is_omap3630() with ISP revision check" [media] dvb: LNA implementation changes [media] v4l2-ioctl: fix W=1 warnings [media] v4l2-ioctl: add blocks check for VIDIOC_SUBDEV_G/S_EDID [media] omap3isp: Fix compilation error in ispreg.h [media] rc-msi-digivox-ii: Add full scan keycodes [media] cx25821: testing the wrong variable [media] tda18271-common: hold the I2C adapter during write transfers [media] ds3000: add module parameter to force firmware upload [media] drivers/media: Remove unnecessary semicolon [media] winbond: remove space from driver name [media] iguanair: cannot send data from the stack [media] omap3isp: Replace cpu_is_omap3630() with ISP revision check [media] dvb-usb: print small buffers via %*ph [media] uvc: Add return code check at vb2_queue_init() [media] em28xx: Replace memcpy with struct assignment [media] bt8xx: Add video4linux control V4L2_CID_COLOR_KILLER [media] mem2mem_testdev: Use devm_kzalloc() in probe ...
2012-10-119P: Fix race between p9_write_work() and p9_fd_request()Simon Derr
Race scenario: thread A thread B p9_write_work() p9_fd_request() if (list_empty (&m->unsent_req_list)) ... spin_lock(&client->lock); req->status = REQ_STATUS_UNSENT; list_add_tail(..., &m->unsent_req_list); spin_unlock(&client->lock); .... if (n & POLLOUT && !test_and_set_bit(Wworksched, &m->wsched) schedule_work(&m->wq); --> not done because Wworksched is set clear_bit(Wworksched, &m->wsched); return; --> nobody will take care of sending the new request. This is not very likely to happen though, because p9_write_work() being called with an empty unsent_req_list is not frequent. But this also means that taking the lock earlier will not be costly. Signed-off-by: Simon Derr <simon.derr@bull.net> Signed-off-by: Eric Van Hensbergen <ericvh@gmail.com>
2012-10-11Merge Trond's bugfixesJ. Bruce Fields
Merge branch 'bugfixes' of git://linux-nfs.org/~trondmy/nfs-2.6 into for-3.7-incoming. Mainly needed for Bryan's "SUNRPC: Set alloc_slot for backchannel tcp ops", without which the 4.1 server oopses.
2012-10-11Update SmPL/Coccinelle section of MAINTAINERSNicolas Palix
This patch updates some email addresses and the new mailing list address. Signed-off-by: Nicolas Palix <nicolas.palix@imag.fr> Acked-by: Julia Lawall <julia.lawall@lip6.fr> Signed-off-by: Michal Marek <mmarek@suse.cz>
2012-10-11coccicheck: Add the rep+ctxt modeNicolas Palix
This adds a 'rep+ctxt' mode which prints the warning message followed by the context. Signed-off-by: Nicolas Palix <nicolas.palix@imag.fr> Signed-off-by: Michal Marek <mmarek@suse.cz>
2012-10-11scripts/coccinelle/tests/odd_ptr_err.cocci: semantic patch for ↵Julia Lawall
IS_ERR/PTR_ERR inconsistency Signed-off-by: Julia Lawall <Julia.Lawall@lip6.fr> Signed-off-by: Michal Marek <mmarek@suse.cz>
2012-10-11Merge tag 'disintegrate-arm64-20121011' of ↵Catalin Marinas
git://git.infradead.org/users/dhowells/linux-headers into uapi UAPI Disintegration 2012-10-11 * tag 'disintegrate-arm64-20121011' of git://git.infradead.org/users/dhowells/linux-headers: UAPI: (Scripted) Disintegrate arch/arm64/include/asm
2012-10-11Merge branch 'tpmdd-next-v3.6' of git://github.com/shpedoikal/linux into ↵James Morris
for-linus
2012-10-11Merge branch 'tpmdd-fixes-v3.6' of git://github.com/shpedoikal/linux into ↵James Morris
for-linus
2012-10-11UAPI: (Scripted) Disintegrate arch/arm64/include/asmDavid Howells
Signed-off-by: David Howells <dhowells@redhat.com> Acked-by: Arnd Bergmann <arnd@arndb.de> Acked-by: Thomas Gleixner <tglx@linutronix.de> Acked-by: Michael Kerrisk <mtk.manpages@gmail.com> Acked-by: Paul E. McKenney <paulmck@linux.vnet.ibm.com> Acked-by: Dave Jones <davej@redhat.com>
2012-10-11Merge branch 'fixes' into for-linusRussell King
Conflicts: arch/arm/kernel/smp.c
2012-10-11arm64: Do not export the compat-specific definitions to the userCatalin Marinas
This patch adds #ifdef __KERNEL__ guards around the COMPAT_* definitions to avoid exporting them to user. AArch32 user requiring the kernel headers must use those generated with ARCH=arm. Signed-off-by: Catalin Marinas <catalin.marinas@arm.com> Acked-by: Will Deacon <will.deacon@arm.com>
2012-10-11arm64: Do not include asm/unistd32.h in asm/unistd.hCatalin Marinas
This patch only includes asm/unistd32.h where necessary and removes its inclusion in the asm/unistd.h file. The __SYSCALL_COMPAT guard is dropped. Signed-off-by: Catalin Marinas <catalin.marinas@arm.com> Acked-by: Will Deacon <will.deacon@arm.com>
2012-10-11arm64: Remove unused definitions from asm/unistd32.hCatalin Marinas
This patch removes the compat __NR_* definitions from the unistd32.h file and only keeps those that are used by the AArch64 kernel with a new __NR_compat_* prefix. The additional wrapper definitions in arch/arm64/kernel/sys32.S have been removed and the actual wrapper names included in the asm/unistd32.h file. Signed-off-by: Catalin Marinas <catalin.marinas@arm.com> Acked-by: Will Deacon <will.deacon@arm.com>
2012-10-11openrisc: mask interrupts in irq_mask_ack functionGong Tao
or1k_pic_mask_ack was failing to actually mask the IRQ. Signed-off-by: Gong Tao <gongtao0607@gmail.com> Signed-off-by: Jonas Bonn <jonas@southpole.se>
2012-10-11openrisc: fix typos in comments and warningsJonas Bonn
Signed-off-by: Jonas Bonn <jonas@southpole.se>
2012-10-11openrisc: PIC should act on domain-local irqsJonas Bonn
Now that IRQ domains are in use, we should be acting on domain-local IRQ numbers (hwirq) instead of 'global' ones. Signed-off-by: Jonas Bonn <jonas@southpole.se>
2012-10-11openrisc: Make cpu_relax() invoke barrier()Vladimir Murzin
Make cpu_relax() invoke barrier() to be the same as other arches. Signed-off-by: Vladimir Murzin <murzin.v@gmail.com> Signed-off-by: Jonas Bonn <jonas@southpole.se>
2012-10-11Merge branch 'next' into for-linusDmitry Torokhov
Prepare second set of updates for 3.7 merge window (Wacom driver update and patches extending number of input minors).
2012-10-11md: refine reporting of resync/reshape delays.NeilBrown
If 'resync_max' is set to 0 (as is often done when starting a reshape, so the mdadm can remain in control during a sensitive period), and if the reshape request is initially delayed because another array using the same array is resyncing or reshaping etc, when user-space cannot easily tell when the delay changes from being due to a conflicting reshape, to being due to resync_max = 0. So introduce a new state: (curr_resync == 3) to reflect this, make sure it is visible both via /proc/mdstat and via the "sync_completed" sysfs attribute, and ensure that the event transition from one delay state to the other is properly notified. Signed-off-by: NeilBrown <neilb@suse.de>
2012-10-11md/raid5: be careful not to resize_stripes too big.NeilBrown
When a RAID5 is reshaping, conf->raid_disks is increased before mddev->delta_disks becomes zero. This can result in check_reshape calling resize_stripes with a number that is too large. This particularly happens when md_check_recovery calls ->check_reshape(). If we use ->previous_raid_disks, we don't risk this. Signed-off-by: NeilBrown <neilb@suse.de>
2012-10-11md: make sure manual changes to recovery checkpoint are saved.NeilBrown
If you make an array bigger but suppress resync of the new region with mdadm --grow /dev/mdX --size=max --assume-clean then stop the array before anything is written to it, the effect of the "--assume-clean" is lost and the array will resync the new space when restarted. So ensure that we update the metadata in the case. Reported-by: Sebastian Riemer <sebastian.riemer@profitbricks.com> Signed-off-by: NeilBrown <neilb@suse.de>
2012-10-11md/raid10: use correct limit variableDan Carpenter
Clang complains that we are assigning a variable to itself. This should be using bad_sectors like the similar earlier check does. Bug has been present since 3.1-rc1. It is minor but could conceivably cause corruption or other bad behaviour. Cc: stable@vger.kernel.org Signed-off-by: Dan Carpenter <dan.carpenter@oracle.com> Signed-off-by: NeilBrown <neilb@suse.de>
2012-10-11md: writing to sync_action should clear the read-auto state.NeilBrown
In some cases array are started in 'read-auto' state where in nothing gets written to any device until the array is written to. The purpose of this is to make accidental auto-assembly of the wrong arrays less of a risk, and to allow arrays to be started to read suspend-to-disk images without actually changing anything (as might happen if the array were dirty and a resync seemed necessary). Explicitly writing the 'sync_action' for a read-auto array currently doesn't clear the read-auto state, so the sync action doesn't happen, which can be confusing. So allow any successful write to sync_action to clear any read-auto state. Reported-by: Alexander Kühn <alexander.kuehn@nagilum.de> Signed-off-by: NeilBrown <neilb@suse.de>
2012-10-11Subject: [PATCH] md:change resync_mismatches to atomic64_t to avoid racesJianpeng Ma
Now that multiple threads can handle stripes, it is safer to use an atomic64_t for resync_mismatches, to avoid update races. Signed-off-by: Jianpeng Ma <majianpeng@gmail.com> Signed-off-by: NeilBrown <neilb@suse.de>
2012-10-10e1000e: Change wthresh to 1 to avoid possible Tx stallsHiroaki SHIMODA
This patch originated from Hiroaki SHIMODA but has been modified by Intel with some minor cleanups and additional commit log text. Denys Fedoryshchenko and others reported Tx stalls on e1000e with BQL enabled. Issue was root caused to hardware delays. They were introduced because some of the e1000e hardware with transmit writeback bursting enabled, waits until the driver does an explict flush OR there are WTHRESH descriptors to write back. Sometimes the delays in question were on the order of seconds, causing visible lag for ssh sessions and unacceptable tx completion latency, especially for BQL enabled kernels. To avoid possible Tx stalls, change WTHRESH back to 1. The current plan is to investigate a method for re-enabling WTHRESH while not harming BQL, but those patches will be later for net-next if they work. please enqueue for stable since v3.3 as this bug was introduced in commit 3f0cfa3bc11e7f00c9994e0f469cbc0e7da7b00c Author: Tom Herbert <therbert@google.com> Date: Mon Nov 28 16:33:16 2011 +0000 e1000e: Support for byte queue limits Changes to e1000e to use byte queue limits. Reported-by: Denys Fedoryshchenko <denys@visp.net.lb> Tested-by: Denys Fedoryshchenko <denys@visp.net.lb> Signed-off-by: Hiroaki SHIMODA <shimoda.hiroaki@gmail.com> CC: eric.dumazet@gmail.com CC: therbert@google.com Signed-off-by: Jesse Brandeburg <jesse.brandeburg@intel.com> Signed-off-by: David S. Miller <davem@davemloft.net>
generated by cgit v1.2.3 (git 2.25.1) at 2024-07-22 22:43:04 +0000