- 29 Nov, 2013 25 commits
-
-
Rashika Kheria authored
commit 46a51c80 upstream. This patch fixes the bug in reset_store caused by accessing NULL pointer. The bdev gets its value from bdget_disk() which could fail when memory pressure is severe and hence can return NULL because allocation of inode in bdget could fail. Hence, this patch introduces a check for bdev to prevent reference to a NULL pointer in the later part of the code. It also removes unnecessary check of bdev for fsync_bdev(). Acked-by:
Jerome Marchand <jmarchan@redhat.com> Signed-off-by:
Rashika Kheria <rashika.kheria@gmail.com> Acked-by:
Minchan Kim <minchan@kernel.org> Signed-off-by:
Greg Kroah-Hartman <gregkh@linuxfoundation.org>
-
Rafael J. Wysocki authored
commit 176a88d7 upstream. According to the ACPI spec (5.0, Section 6.3.5), the "Device insertion in progress (pending)" (0x80) _OST status code is reserved for the "Insertion Processing" (0x200) source event which is "a result of an OSPM action". Specifically, it is not a notification, so that status code should not be used during notification processing, which unfortunately is done by acpi_scan_bus_device_check(). For this reason, drop the ACPI_OST_SC_INSERT_IN_PROGRESS _OST status evaluation from there (it was a mistake to put it in there in the first place). Signed-off-by:
Rafael J. Wysocki <rafael.j.wysocki@intel.com> Signed-off-by:
-
Rafael J. Wysocki authored
commit 2441191a upstream. It is required to do get_device() on the struct acpi_device in question before passing it to acpi_bus_hot_remove_device() through acpi_os_hotplug_execute(), because acpi_bus_hot_remove_device() calls acpi_scan_hot_remove() that does put_device() on that object. The ACPI PCI root removal routine, handle_root_bridge_removal(), doesn't do that, which may lead to premature freeing of the device object or to executing put_device() on an object that has been freed already. Fix this problem by making handle_root_bridge_removal() use get_device() as appropriate. Signed-off-by:
Toshi Kani <toshi.kani@hp.com> Signed-off-by:
-
Aaron Lu authored
commit 2c62333a upstream. Some firmware doesn't initialize initial backlight level to a proper value and _BQC will return 0 on first time evaluation. We used to be able to detect such incorrect value with our code logic, as value 0 normally isn't a valid value in _BCL. But with the introduction of Win8, firmware begins to fill _BCL with values from 0 to 100, now 0 becomes a valid value but that value will make user's screen black. This patch test initial _BQC for value 0, if such a value is returned, do not use it. References: https://bugzilla.kernel.org/show_bug.cgi?id=64031 References: https://bugzilla.kernel.org/show_bug.cgi?id=61231 References: https://bugzilla.kernel.org/show_bug.cgi?id=63111Reported-by:
Qingshuai Tian <qingshuai.tian@intel.com> Tested-by: Aaron Lu <aaron.lu@intel.com> # on "Idealpad u330p" Reported-and-tested-by: <erno@iki.fi> # on "Acer Aspire V5-573G" Reported-and-tested-by: Kirill Tkhai <tkhai@yandex.ru> # on "HP 250 G1" Signed-off-by:
Aaron Lu <aaron.lu@intel.com> Signed-off-by:
-
Puneet Kumar authored
commit 36b15875 upstream. A bug was introduced by commit b76b51ba ('ACPI / EC: Add more debug info and trivial code cleanup') that erroneously caused the struct member to be accessed before acquiring the required lock. This change fixes it by ensuring the lock acquisition is done first. Found by Aaron Durbin <adurbin@chromium.org> Fixes: b76b51ba ('ACPI / EC: Add more debug info and trivial code cleanup') References: http://crbug.com/319019Signed-off-by:
Puneet Kumar <puneetster@chromium.org> Reviewed-by:
Aaron Durbin <adurbin@chromium.org> [olof: Commit message reworded a bit] Signed-off-by:
Olof Johansson <olof@lixom.net> Signed-off-by:
-
Steven Rostedt authored
commit 12ae030d upstream. The current default perf paranoid level is "1" which has "perf_paranoid_kernel()" return false, and giving any operations that use it, access to normal users. Unfortunately, this includes function tracing and normal users should not be allowed to enable function tracing by default. The proper level is defined at "-1" (full perf access), which "perf_paranoid_tracepoint_raw()" will only give access to. Use that check instead for enabling function tracing. Reported-by:
Dave Jones <davej@redhat.com> Reported-by:
Vince Weaver <vincent.weaver@maine.edu> Tested-by:
Steven Rostedt <rostedt@goodmis.org> Signed-off-by:
-
Peter Zijlstra authored
commit ea811747 upstream. Mike reported that commit 7d1a9417 ("x86: Use generic idle loop") regressed several workloads and caused excessive reschedule interrupts. The patch in question failed to notice that the x86 code had an inverted sense of the polling state versus the new generic code (x86: default polling, generic: default !polling). Fix the two prominent x86 mwait based idle drivers and introduce a few new generic polling helpers (fixing the wrong smp_mb__after_clear_bit usage). Also switch the idle routines to using tif_need_resched() which is an immediate TIF_NEED_RESCHED test as opposed to need_resched which will end up being slightly different. Reported-by:
Mike Galbraith <bitbucket@online.de> Signed-off-by:
Peter Zijlstra <peterz@infradead.org> Cc: lenb@kernel.org Cc: tglx@linutronix.de Link: http://lkml.kernel.org/n/tip-nc03imb0etuefmzybzj7sprf@git.kernel.orgSigned-off-by:
Ingo Molnar <mingo@kernel.org> Signed-off-by:
-
Suman Anna authored
commit 6423d6df upstream. The hwmod init sequence involves initializing and idling all the hwmods during bootup. If a module class has sysconfig, the init sequence utilizes the module register base for performing any sysc configuration. The module address space is being removed from hwmod database and retrieved from the <reg> property of the corresponding DT node. If a hwmod does not have its corresponding DT node defined and the memory address space is not defined in the corresponding omap_hwmod_ocp_if, then the module register target address space would be NULL and any sysc programming would result in a NULL pointer dereference and a kernel boot hang. Handle this scenario by checking for a valid module address space during the _init of each hwmod, and leaving it in the registered state if no module register address base is defined in either of the hwmod data or the DT data. Signed-off-by:
Suman Anna <s-anna@ti.com> Acked-by:
Santosh Shilimkar <santosh.shilimkar@ti.com> Tested-by:
Nishanth Menon <nm@ti.com> Acked-by:
Tony Lindgren <tony@atomide.com> [paul@pwsan.com: use -ENXIO rather than -ENOMEM to indicate a missing address space error; fixed checkpatch.pl problem] Signed-off-by:
Paul Walmsley <paul@pwsan.com> Signed-off-by:
-
Takashi Iwai authored
commit cab66613 upstream. SNDRV_CARDS can be specified via Kconfig since 3.11 kernel, so this can be over 32bit integer range, which leads to a build error. Signed-off-by:
Takashi Iwai <tiwai@suse.de> Signed-off-by:
Matthew Garrett <matthew.garrett@nebula.com> Signed-off-by:
-
Ben Skeggs authored
commit 13d2b35a upstream. Signed-off-by:
Ben Skeggs <bskeggs@redhat.com> Cc: Ilia Mirkin <imirkin@alum.mit.edu> Signed-off-by:
-
John Stultz authored
commit 59573240 upstream. Hopefully this isn't too late for 3.12. In commit 7dc19d5a (convert shrinkers to new count/scan API) the return value to PURGE_ALL_CACHES was dropped, causing -EPERM to always be returned. This patch re-adds the ret assignment, setting it to the the ashmem_shrink_count(), which is the lru_count. (Sorry this was missed in the review!) Fixes: 7dc19d5a ("convert shrinkers to new count/scan API") Cc: Colin Cross <ccross@android.com> Cc: Android Kernel Team <kernel-team@android.com> Cc: Glauber Costa <glommer@openvz.org> Reported-by:
YongQin Liu <yongqin.liu@linaro.org> Signed-off-by:
John Stultz <john.stultz@linaro.org> Acked-by:
Dave Chinner <dchinner@redhat.com> Signed-off-by:
-
Laurent Pinchart authored
commit 0a5a5499 upstream. The driver registers a backlight device and thus requires BACKLIGHT_CLASS_DEVICE to be selected to avoid compilation breakages. Reported-by:
Russell King <linux@arm.linux.org.uk> Signed-off-by:
Laurent Pinchart <laurent.pinchart@ideasonboard.com> Signed-off-by:
Dave Airlie <airlied@redhat.com> Signed-off-by:
-
Greg Thelen authored
commit a399b29d upstream. When IPC_RMID races with other shm operations there's potential for use-after-free of the shm object's associated file (shm_file). Here's the race before this patch: TASK 1 TASK 2 ------ ------ shm_rmid() ipc_lock_object() shmctl() shp = shm_obtain_object_check() shm_destroy() shum_unlock() fput(shp->shm_file) ipc_lock_object() shmem_lock(shp->shm_file) <OOPS> The oops is caused because shm_destroy() calls fput() after dropping the ipc_lock. fput() clears the file's f_inode, f_path.dentry, and f_path.mnt, which causes various NULL pointer references in task 2. I reliably see the oops in task 2 if with shmlock, shmu This patch fixes the races by: 1) set shm_file=NULL in shm_destroy() while holding ipc_object_lock(). 2) modify at risk operations to check shm_file while holding ipc_object_lock(). Example workloads, which each trigger oops... Workload 1: while true; do id=$(shmget 1 4096) shm_rmid $id & shmlock $id & wait done The oops stack shows accessing NULL f_inode due to racing fput: _raw_spin_lock shmem_lock SyS_shmctl Workload 2: while true; do id=$(shmget 1 4096) shmat $id 4096 & shm_rmid $id & wait done The oops stack is similar to workload 1 due to NULL f_inode: touch_atime shmem_mmap shm_mmap mmap_region do_mmap_pgoff do_shmat SyS_shmat Workload 3: while true; do id=$(shmget 1 4096) shmlock $id shm_rmid $id & shmunlock $id & wait done The oops stack shows second fput tripping on an NULL f_inode. The first fput() completed via from shm_destroy(), but a racing thread did a get_file() and queued this fput(): locks_remove_flock __fput ____fput task_work_run do_notify_resume int_signal Fixes: c2c737a0 ("ipc,shm: shorten critical region for shmat") Fixes: 2caacaa8 ("ipc,shm: shorten critical region for shmctl") Signed-off-by:
Greg Thelen <gthelen@google.com> Cc: Davidlohr Bueso <davidlohr@hp.com> Cc: Rik van Riel <riel@redhat.com> Cc: Manfred Spraul <manfred@colorfullife.com> Signed-off-by:
Andrew Morton <akpm@linux-foundation.org> Signed-off-by:
Linus Torvalds <torvalds@linux-foundation.org> Signed-off-by:
-
Jesper Nilsson authored
commit 3a72660b upstream. Commit 2caacaa8 ("ipc,shm: shorten critical region for shmctl") restructured the ipc shm to shorten critical region, but introduced a path where the return value could be -EPERM, even if the operation actually was performed. Before the commit, the err return value was reset by the return value from security_shm_shmctl() after the if (!ns_capable(...)) statement. Now, we still exit the if statement with err set to -EPERM, and in the case of SHM_UNLOCK, it is not reset at all, and used as the return value from shmctl. To fix this, we only set err when errors occur, leaving the fallthrough case alone. Signed-off-by:
Jesper Nilsson <jesper.nilsson@axis.com> Cc: Davidlohr Bueso <davidlohr@hp.com> Cc: Rik van Riel <riel@redhat.com> Cc: Michel Lespinasse <walken@google.com> Cc: Al Viro <viro@zeniv.linux.org.uk> Signed-off-by:
-
hahnjo authored
commit b54629e2 upstream. This fixes bug 62491 (https://bugzilla.kernel.org/show_bug.cgi?id=62491). After resuming some users got the following error flooding the kernel log: alx 0000:02:00.0: invalid PHY speed/duplex: 0xffff Signed-off-by:
Jonas Hahnfeld <linux@hahnjo.de> Signed-off-by:
David S. Miller <davem@davemloft.net> Cc: hahnjo <linux@hahnjo.de> Signed-off-by:
-
Markus Pargmann authored
commit 5d0f801a upstream. If we handle end of block messages with higher priority than a lost message, we can run into an endless interrupt loop. This is reproducable with a am335x processor and "cansequence -r" at 1Mbit. As soon as we loose a packet we can't escape from an interrupt loop. This patch fixes the problem by handling lost packets before EOB packets. Signed-off-by:
Markus Pargmann <mpa@pengutronix.de> Signed-off-by:
Marc Kleine-Budde <mkl@pengutronix.de> Signed-off-by:
-
Herbert Xu authored
commit f262f0f5 upstream. The cbc-aes-s390 algorithm incorrectly places the IV in the tfm data structure. As the tfm is shared between multiple threads, this introduces a possibility of data corruption. This patch fixes this by moving the parameter block containing the IV and key onto the stack (the block is 48 bytes long). The same bug exists elsewhere in the s390 crypto system and they will be fixed in subsequent patches. Signed-off-by:
Herbert Xu <herbert@gondor.apana.org.au> Signed-off-by:
-
Neil Horman authored
commit 714b33d1 upstream. Stephan Mueller reported to me recently a error in random number generation in the ansi cprng. If several small requests are made that are less than the instances block size, the remainder for loop code doesn't increment rand_data_valid in the last iteration, meaning that the last bytes in the rand_data buffer gets reused on the subsequent smaller-than-a-block request for random data. The fix is pretty easy, just re-code the for loop to make sure that rand_data_valid gets incremented appropriately Signed-off-by:
Neil Horman <nhorman@tuxdriver.com> Reported-by:
Stephan Mueller <stephan.mueller@atsec.com> CC: Stephan Mueller <stephan.mueller@atsec.com> CC: Petr Matousek <pmatouse@redhat.com> CC: Herbert Xu <herbert@gondor.apana.org.au> CC: "David S. Miller" <davem@davemloft.net> Signed-off-by:
-
Josef Bacik authored
commit 4577b014 upstream. A user reported a problem where they were getting csum errors when running a balance and running systemd's journal. This is because systemd is awesome and fallocate()'s its log space and writes into it. Unfortunately we assume that when we read in all the csums for an extent that they are sequential starting at the bytenr we care about. This obviously isn't the case for prealloc extents, where we could have written to the middle of the prealloc extent only, which means the csum would be for the bytenr in the middle of our range and not the front of our range. Fix this by offsetting the new bytenr we are logging to based on the original bytenr the csum was for. With this patch I no longer see the csum errors I was seeing. Thanks, Reported-by:
Chris Murphy <lists@colorremedies.com> Signed-off-by:
Josef Bacik <jbacik@fusionio.com> Signed-off-by:
Chris Mason <chris.mason@fusionio.com> Signed-off-by:
-
Olivier Sobrie authored
commit 896e23bd upstream. Some devices, like the Kvaser Memorator Professional, have several bulk in endpoints. Only the first one found must be used by the driver. The same holds for the bulk out endpoint. The official Kvaser driver (leaf) was used as reference for this patch. Signed-off-by:
Olivier Sobrie <olivier@sobrie.be> Signed-off-by:
-
Johan Hovold authored
commit a91ccd26 upstream. Make sure to return errors from tiocmget rather than rely on uninitialised stack data. Signed-off-by:
Johan Hovold <jhovold@gmail.com> Signed-off-by:
-
Eric Sandeen authored
commit 10e6e65d upstream. Today, if xfs_sb_read_verify encounters a v4 superblock with junk past v4 fields which includes data in sb_crc, it will be treated as a failing checksum and a significant corruption. There are known prior bugs which leave junk at the end of the V4 superblock; we don't need to actually fail the verification in this case if other checks pan out ok. So if this is a secondary superblock, and the primary superblock doesn't indicate that this is a V5 filesystem, don't treat this as an actual checksum failure. We should probably check the garbage condition as we do in xfs_repair, and possibly warn about it or self-heal, but that's a different scope of work. Stable folks: This can go back to v3.10, which is what introduced the sb CRC checking that is tripped up by old, stale, incorrect V4 superblocks w/ unzeroed bits. Signed-off-by:
Eric Sandeen <sandeen@redhat.com> Acked-by:
Dave Chinner <david@fromorbit.com> Reviewed-by:
Mark Tinguely <tinguely@sgi.com> Signed-off-by:
Ben Myers <bpm@sgi.com> Signed-off-by:
-
Mahesh Rajashekhara authored
commit b4789b8e upstream. It appears that driver runs into a problem here if fibsize is too small because we allocate user_srbcmd with fibsize size only but later we access it until user_srbcmd->sg.count to copy it over to srbcmd. It is not correct to test (fibsize < sizeof(*user_srbcmd)) because this structure already includes one sg element and this is not needed for commands without data. So, we would recommend to add the following (instead of test for fibsize == 0). Signed-off-by:
Mahesh Rajashekhara <Mahesh.Rajashekhara@pmcs.com> Reported-by:
Nico Golde <nico@ngolde.de> Reported-by:
Fabian Yamaguchi <fabs@goesec.de> Signed-off-by:
-
Dan Carpenter authored
commit a497e47d upstream. If we do a zero size allocation then it will oops. Also we can't be sure the user passes us a NUL terminated string so I've added a terminator. This code can only be triggered by root. Reported-by:
Dan Carpenter <dan.carpenter@oracle.com> Acked-by:
Dan Williams <dcbw@redhat.com> Signed-off-by:
John W. Linville <linville@tuxdriver.com> Cc: Kees Cook <keescook@chromium.org> Signed-off-by:
-
Al Viro authored
commit ede4cebc upstream. ... and equivalent is needed in 3.12; it's broken there as well Signed-off-by:
Al Viro <viro@zeniv.linux.org.uk> Reported-by:
Michael Marineau <michael.marineau@coreos.com> Tested-by:
Waiman Long <Waiman.Long@hp.com> Signed-off-by:
-
- 20 Nov, 2013 15 commits
-
-
Greg Kroah-Hartman authored
-
Xenia Ragiadakou authored
commit 9df89d85 upstream. This patch sets the lpm_capable field for root hubs with LPM capabilities. Signed-off-by:
Xenia Ragiadakou <burzalodowa@gmail.com> Reported-by:
Martin MOKREJS <mmokrejs@gmail.com> Suggested-by:
Alan Stern <stern@rowland.harvard.edu> Signed-off-by:
-
Johan Hovold authored
commit 9d3fde86 upstream. Move probe out of __init section and don't use platform_driver_probe which cannot be used with deferred probing. Since commit e9354576 ("gpiolib: Defer failed gpio requests by default") this driver might return -EPROBE_DEFER if a gpio_request fails. Cc: Richard Purdie <rpurdie@rpsys.net> Cc: Jingoo Han <jg1.han@samsung.com> Cc: Jean-Christophe Plagniol-Villard <plagnioj@jcrosoft.com> Signed-off-by:
-
Johan Hovold authored
commit 5c6d6fd1 upstream. Two drivers (atmel-pwm-bl and leds-atmel-pwm) currently depend on the atmel_pwm driver to have bound to any pwm-device before their devices are probed. Support deferred probing of such devices by making sure to return -EPROBE_DEFER from pwm_channel_alloc when no pwm-device has yet been bound. Signed-off-by:
-
Steven Rostedt authored
commit 057db848 upstream. Andrey reported the following report: ERROR: AddressSanitizer: heap-buffer-overflow on address ffff8800359c99f3 ffff8800359c99f3 is located 0 bytes to the right of 243-byte region [ffff8800359c9900, ffff8800359c99f3) Accessed by thread T13003: #0 ffffffff810dd2da (asan_report_error+0x32a/0x440) #1 ffffffff810dc6b0 (asan_check_region+0x30/0x40) #2 ffffffff810dd4d3 (__tsan_write1+0x13/0x20) #3 ffffffff811cd19e (ftrace_regex_release+0x1be/0x260) #4 ffffffff812a1065 (__fput+0x155/0x360) #5 ffffffff812a12de (____fput+0x1e/0x30) #6 ffffffff8111708d (task_work_run+0x10d/0x140) #7 ffffffff810ea043 (do_exit+0x433/0x11f0) #8 ffffffff810eaee4 (do_group_exit+0x84/0x130) #9 ffffffff810eafb1 (SyS_exit_group+0x21/0x30) #10 ffffffff81928782 (system_call_fastpath+0x16/0x1b) Allocated by thread T5167: #0 ffffffff810dc778 (asan_slab_alloc+0x48/0xc0) #1 ffffffff8128337c (__kmalloc+0xbc/0x500) #2 ffffffff811d9d54 (trace_parser_get_init+0x34/0x90) #3 ffffffff811cd7b3 (ftrace_regex_open+0x83/0x2e0) #4 ffffffff811cda7d (ftrace_filter_open+0x2d/0x40) #5 ffffffff8129b4ff (do_dentry_open+0x32f/0x430) #6 ffffffff8129b668 (finish_open+0x68/0xa0) #7 ffffffff812b66ac (do_last+0xb8c/0x1710) #8 ffffffff812b7350 (path_openat+0x120/0xb50) #9 ffffffff812b8884 (do_filp_open+0x54/0xb0) #10 ffffffff8129d36c (do_sys_open+0x1ac/0x2c0) #11 ffffffff8129d4b7 (SyS_open+0x37/0x50) #12 ffffffff81928782 (system_call_fastpath+0x16/0x1b) Shadow bytes around the buggy address: ffff8800359c9700: fd fd fd fd fd fd fd fd fd fd fd fd fd fd fd fd ffff8800359c9780: fd fd fd fd fd fd fd fd fa fa fa fa fa fa fa fa ffff8800359c9800: fa fa fa fa fa fa fa fa fa fa fa fa fa fa fa fa ffff8800359c9880: fa fa fa fa fa fa fa fa fa fa fa fa fa fa fa fa ffff8800359c9900: 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 =>ffff8800359c9980: 00 00 00 00 00 00 00 00 00 00 00 00 00 00[03]fb ffff8800359c9a00: fa fa fa fa fa fa fa fa fa fa fa fa fa fa fa fa ffff8800359c9a80: fa fa fa fa fa fa fa fa fa fa fa fa fa fa fa fa ffff8800359c9b00: fa fa fa fa fa fa fa fa 00 00 00 00 00 00 00 00 ffff8800359c9b80: 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 ffff8800359c9c00: 00 00 00 00 00 00 00 00 fa fa fa fa fa fa fa fa Shadow byte legend (one shadow byte represents 8 application bytes): Addressable: 00 Partially addressable: 01 02 03 04 05 06 07 Heap redzone: fa Heap kmalloc redzone: fb Freed heap region: fd Shadow gap: fe The out-of-bounds access happens on 'parser->buffer[parser->idx] = 0;' Although the crash happened in ftrace_regex_open() the real bug occurred in trace_get_user() where there's an incrementation to parser->idx without a check against the size. The way it is triggered is if userspace sends in 128 characters (EVENT_BUF_SIZE + 1), the loop that reads the last character stores it and then breaks out because there is no more characters. Then the last character is read to determine what to do next, and the index is incremented without checking size. Then the caller of trace_get_user() usually nulls out the last character with a zero, but since the index is equal to the size, it writes a nul character after the allocated space, which can corrupt memory. Luckily, only root user has write access to this file. Link: http://lkml.kernel.org/r/20131009222323.04fd1a0d@gandalf.local.homeReported-by:
Andrey Konovalov <andreyknvl@google.com> Signed-off-by:
-
Anssi Hannula authored
commit 56cac413 upstream. hdmi_setup_fake_chmap() is supposed to set the reported channel map when the channel map is not specified by the user. However, the function indexes channel_allocations[] with a wrong value and extracts the wrong nibble from hdmi_channel_mapping[], causing wrong channel maps to be shown. Fix those issues. Tested on Intel HDMI to correctly generate various channel maps, for example 3,4,14,15,7,8,5,6 (instead of incorrect 3,4,8,7,5,6,14,0) for standard 7.1 channel audio. (Note that the side and rear channels are reported as RL/RR and RLC/RRC, respectively, as per the CEA-861 standard, instead of the more traditional SL/SR and RL/RR.) Note that this only fixes the layouts that only contain traditional 7.1 speakers (2.0, 2.1, 4.0, 5.1, 7.1, etc.). E.g. the rear center of 6.1 is still being shown wrongly due to an issue with from_cea_slot() which will be fixed in a later patch. Signed-off-by:
Anssi Hannula <anssi.hannula@iki.fi> Signed-off-by:
-
Rui li authored
commit 0636fc50 upstream. Signed-off-by:
Rui li <li.rui27@zte.com.cn> Cc: stable <stable@vger.kernel.org> Signed-off-by:
-
Gerd Hoffmann authored
commit 7ad96847 upstream. This patch adds a pci stub driver to hyper-fb. The hyperv framebuffer driver will bind to the pci device then, so linux kernel and userspace know there is a proper kernel driver for the device active. lspci shows this for example: [root@dhcp231 ~]# lspci -vs8 00:08.0 VGA compatible controller: Microsoft Corporation Hyper-V virtual VGA (prog-if 00 [VGA controller]) Flags: bus master, fast devsel, latency 0, IRQ 11 Memory at f8000000 (32-bit, non-prefetchable) [size=64M] Expansion ROM at <unassigned> [disabled] Kernel driver in use: hyperv_fb Another effect is that the xorg vesa driver will not attach to the device and thus the Xorg server will automatically use the fbdev driver instead. Signed-off-by:
Gerd Hoffmann <kraxel@redhat.com> Acked-by:
Haiyang Zhang <haiyangz@microsoft.com> Signed-off-by:
Tomi Valkeinen <tomi.valkeinen@ti.com> Signed-off-by:
-
Hannes Frederic Sowa authored
[ Upstream commit 01ba16d6 ] On receiving a packet too big icmp error we update the expire value by calling rt6_update_expires. This function uses dst_set_expires which is implemented that it can only reduce the expiration value of the dst entry. If we insert new routing non-expiry information into the ipv6 fib where we already have a matching rt6_info we only clear the RTF_EXPIRES flag in rt6i_flags and leave the dst.expires value as is. When new mtu information arrives for that cached dst_entry we again call dst_set_expires. This time it won't update the dst.expire value because we left the dst.expire value intact from the last update. So dst_set_expires won't touch dst.expires. Fix this by resetting dst.expires when clearing the RTF_EXPIRE flag. dst_set_expires checks for a zero expiration and updates the dst.expires. In the past this (not updating dst.expires) was necessary because dst.expire was placed in a union with the dst_entry *from reference and rt6_clean_expires did assign NULL to it. This split happend in ecd98837 ("ipv6: fix race condition regarding dst->expires and dst->from"). Reported-by:
Steinar H. Gunderson <sgunderson@bigfoot.com> Reported-by:
Valentijn Sessink <valentyn@blub.net> Cc: YOSHIFUJI Hideaki <yoshfuji@linux-ipv6.org> Acked-by:
Eric Dumazet <edumazet@google.com> Tested-by:
Hannes Frederic Sowa <hannes@stressinduktion.org> Signed-off-by:
-
Hannes Frederic Sowa authored
[ Upstream commit e3bc10bd ] On receiving a packet too big icmp error we check if our current cached dst_entry in the socket is still valid. This validation check did not care about the expiration of the (cached) route. The error path I traced down: The socket receives a packet too big mtu notification. It still has a valid dst_entry and thus issues the ip6_rt_pmtu_update on this dst_entry, setting RTF_EXPIRE and updates the dst.expiration value (which could fail because of not up-to-date expiration values, see previous patch). In some seldom cases we race with a) the ip6_fib gc or b) another routing lookup which would result in a recreation of the cached rt6_info from its parent non-cached rt6_info. While copying the rt6_info we reinitialize the metrics store by copying it over from the parent thus invalidating the just installed pmtu update (both dsts use the same key to the inetpeer storage). The dst_entry with the just invalidated metrics data would just get its RTF_EXPIRES flag cleared and would continue to stay valid for the socket. We should have not issued the pmtu update on the already expired dst_entry in the first placed. By checking the expiration on the dst entry and doing a relookup in case it is out of date we close the race because we would install a new rt6_info into the fib before we issue the pmtu update, thus closing this race. Not reliably updating the dst.expire value was fixed by the patch "ipv6: reset dst.expires value when clearing expire flag". Reported-by:
-
Yuchung Cheng authored
[ Upstream commit 2f715c1d ] Patch ed08495c "tcp: use RTT from SACK for RTO" always re-arms RTO upon obtaining a RTT sample from newly sacked data. But technically RTO should only be re-armed when the data sent before the last (re)transmission of write queue head are (s)acked. Otherwise the RTO may continue to extend during loss recovery on data sent in the future. Note that RTTs from ACK or timestamps do not have this problem, as the RTT source must be from data sent before. The new RTO re-arm policy is 1) Always re-arm RTO if SND.UNA is advanced 2) Re-arm RTO if sack RTT is available, provided the sacked data was sent before the last time write_queue_head was sent. Signed-off-by:
Larry Brakmo <brakmo@google.com> Signed-off-by:
Yuchung Cheng <ycheng@google.com> Acked-by:
Neal Cardwell <ncardwell@google.com> Acked-by:
-
Yuchung Cheng authored
[ Upstream commit 2909d874 ] Patch ed08495c "tcp: use RTT from SACK for RTO" has a bug that it does not check if the ACK acknowledge new data before taking the RTT sample from TCP timestamps. This patch adds the check back as required by the RFC. Signed-off-by:
-
Yuchung Cheng authored
[ Upstream commit bc15afa3 ] tp->lsndtime may not always be the SYNACK timestamp if a passive Fast Open socket sends data before handshake completes. And if the remote acknowledges both the data and the SYNACK, the RTT sample is already taken in tcp_ack(), so no need to call tcp_update_ack_rtt() in tcp_synack_rtt_meas() aagain. Signed-off-by:
-
Eric Dumazet authored
[ Upstream commit 0d08c42c ] commit 6ff50cd5 ("tcp: gso: do not generate out of order packets") had an heuristic that can trigger a warning in skb_try_coalesce(), because skb->truesize of the gso segments were exactly set to mss. This breaks the requirement that skb->truesize >= skb->len + truesizeof(struct sk_buff); It can trivially be reproduced by : ifconfig lo mtu 1500 ethtool -K lo tso off netperf As the skbs are looped into the TCP networking stack, skb_try_coalesce() warns us of these skb under-estimating their truesize. Signed-off-by:
Alexei Starovoitov <ast@plumgrid.com> Signed-off-by:
-
Ben Hutchings authored
[ Upstream commit 262e827f ] The length calculation here is now invalid on 32-bit architectures, since sk_buff::tail is a pointer and sk_buff::transport_header is an integer offset: drivers/net/ethernet/chelsio/cxgb3/sge.c: In function 'write_ofld_wr': drivers/net/ethernet/chelsio/cxgb3/sge.c:1603:9: warning: passing argument 4 of 'make_sgl' makes integer from pointer without a cast [enabled by default] adap->pdev); ^ drivers/net/ethernet/chelsio/cxgb3/sge.c:964:28: note: expected 'unsigned int' but argument is of type 'sk_buff_data_t' static inline unsigned int make_sgl(const struct sk_buff *skb, ^ Use the appropriate skb accessor functions. Compile-tested only. Signed-off-by:
Ben Hutchings <ben@decadent.org.uk> Fixes: 1a37e412 ('net: Use 16bits for *_headers fields of struct skbuff') Signed-off-by:
-
