kernel/linux.git/drivers/scsi/sg.c, branch v6.19.11

scsi: sg: Fix occasional bogus elapsed time that exceeds timeout

2025-12-17T02:24:28+00:00

A race condition was found in sg_proc_debug_helper(). It was observed on a system using an IBM LTO-9 SAS Tape Drive (ULTRIUM-TD9) and monitoring /proc/scsi/sg/debug every second. A very large elapsed time would sometimes appear. This is caused by two race conditions. We reproduced the issue with an IBM ULTRIUM-HH9 tape drive on an x86_64 architecture. A patched kernel was built, and the race condition could not be observed anymore after the application of this patch. A reproducer C program utilising the scsi_debug module was also built by Changhui Zhong and can be viewed here: https://github.com/MichaelRabek/linux-tests/blob/master/drivers/scsi/sg/sg_race_trigger.c The first race happens between the reading of hp->duration in sg_proc_debug_helper() and request completion in sg_rq_end_io(). The hp->duration member variable may hold either of two types of information: #1 - The start time of the request. This value is present while the request is not yet finished. #2 - The total execution time of the request (end_time - start_time). If sg_proc_debug_helper() executes *after* the value of hp->duration was changed from #1 to #2, but *before* srp->done is set to 1 in sg_rq_end_io(), a fresh timestamp is taken in the else branch, and the elapsed time (value type #2) is subtracted from a timestamp, which cannot yield a valid elapsed time (which is a type #2 value as well). To fix this issue, the value of hp->duration must change under the protection of the sfp->rq_list_lock in sg_rq_end_io(). Since sg_proc_debug_helper() takes this read lock, the change to srp->done and srp->header.duration will happen atomically from the perspective of sg_proc_debug_helper() and the race condition is thus eliminated. The second race condition happens between sg_proc_debug_helper() and sg_new_write(). Even though hp->duration is set to the current time stamp in sg_add_request() under the write lock's protection, it gets overwritten by a call to get_sg_io_hdr(), which calls copy_from_user() to copy struct sg_io_hdr from userspace into kernel space. hp->duration is set to the start time again in sg_common_write(). If sg_proc_debug_helper() is called between these two calls, an arbitrary value set by userspace (usually zero) is used to compute the elapsed time. To fix this issue, hp->duration must be set to the current timestamp again after get_sg_io_hdr() returns successfully. A small race window still exists between get_sg_io_hdr() and setting hp->duration, but this window is only a few instructions wide and does not result in observable issues in practice, as confirmed by testing. Additionally, we fix the format specifier from %d to %u for printing unsigned int values in sg_proc_debug_helper(). Signed-off-by: Michal Rábek Suggested-by: Tomas Henzl Tested-by: Changhui Zhong Reviewed-by: Ewan D. Milne Reviewed-by: John Meneghini Reviewed-by: Tomas Henzl Link: https://patch.msgid.link/20251212160900.64924-1-mrabek@redhat.com Signed-off-by: Martin K. Petersen

scsi: sg: Do not sleep in atomic context

2025-11-20T03:46:36+00:00

sg_finish_rem_req() calls blk_rq_unmap_user(). The latter function may sleep. Hence, call sg_finish_rem_req() with interrupts enabled instead of disabled. Reported-by: syzbot+c01f8e6e73f20459912e@syzkaller.appspotmail.com Closes: https://lore.kernel.org/linux-scsi/691560c4.a70a0220.3124cb.001a.GAE@google.com/ Cc: Hannes Reinecke Cc: stable@vger.kernel.org Fixes: 97d27b0dd015 ("scsi: sg: close race condition in sg_remove_sfp_usercontext()") Signed-off-by: Bart Van Assche Link: https://patch.msgid.link/20251113181643.1108973-1-bvanassche@acm.org Signed-off-by: Martin K. Petersen

scsi: sg: drop nth_page() usage within SG entry

2025-09-21T21:22:08+00:00

It's no longer required to use nth_page() when iterating pages within a single SG entry, so let's drop the nth_page() usage. Link: https://lkml.kernel.org/r/20250901150359.867252-32-david@redhat.com Signed-off-by: David Hildenbrand Reviewed-by: Bart Van Assche Reviewed-by: Lorenzo Stoakes Reviewed-by: Martin K. Petersen Cc: Doug Gilbert Cc: "James E.J. Bottomley" Signed-off-by: Andrew Morton

scsi: sg: Remove unnecessary NULL check before unregister_sysctl_table()

2025-05-21T01:59:44+00:00

unregister_sysctl_table() checks for NULL pointers internally. Remove unneeded NULL check here. Signed-off-by: Chen Ni Link: https://lore.kernel.org/r/20250514032845.2317700-1-nichen@iscas.ac.cn Reviewed-by: John Garry Signed-off-by: Martin K. Petersen

treewide: const qualify ctl_tables where applicable

2025-01-28T12:48:37+00:00

Add the const qualifier to all the ctl_tables in the tree except for watchdog_hardlockup_sysctl, memory_allocation_profiling_sysctls, loadpin_sysctl_table and the ones calling register_net_sysctl (./net, drivers/inifiniband dirs). These are special cases as they use a registration function with a non-const qualified ctl_table argument or modify the arrays before passing them on to the registration function. Constifying ctl_table structs will prevent the modification of proc_handler function pointers as the arrays would reside in .rodata. This is made possible after commit 78eb4ea25cd5 ("sysctl: treewide: constify the ctl_table argument of proc_handlers") constified all the proc_handlers. Created this by running an spatch followed by a sed command: Spatch: virtual patch @ depends on !(file in "net") disable optional_qualifier @ identifier table_name != { watchdog_hardlockup_sysctl, iwcm_ctl_table, ucma_ctl_table, memory_allocation_profiling_sysctls, loadpin_sysctl_table }; @@ + const struct ctl_table table_name [] = { ... }; sed: sed --in-place \ -e "s/struct ctl_table .table = &uts_kern/const struct ctl_table *table = \&uts_kern/" \ kernel/utsname_sysctl.c Reviewed-by: Song Liu Acked-by: Steven Rostedt (Google) # for kernel/trace/ Reviewed-by: Martin K. Petersen # SCSI Reviewed-by: Darrick J. Wong # xfs Acked-by: Jani Nikula Acked-by: Corey Minyard Acked-by: Wei Liu Acked-by: Thomas Gleixner Reviewed-by: Bill O'Donnell Acked-by: Baoquan He Acked-by: Ashutosh Dixit Acked-by: Anna Schumaker Signed-off-by: Joel Granados

scsi: sg: Fix slab-use-after-free read in sg_release()

2024-12-04T18:22:59+00:00

Fix a use-after-free bug in sg_release(), detected by syzbot with KASAN: BUG: KASAN: slab-use-after-free in lock_release+0x151/0xa30 kernel/locking/lockdep.c:5838 __mutex_unlock_slowpath+0xe2/0x750 kernel/locking/mutex.c:912 sg_release+0x1f4/0x2e0 drivers/scsi/sg.c:407 In sg_release(), the function kref_put(&sfp->f_ref, sg_remove_sfp) is called before releasing the open_rel_lock mutex. The kref_put() call may decrement the reference count of sfp to zero, triggering its cleanup through sg_remove_sfp(). This cleanup includes scheduling deferred work via sg_remove_sfp_usercontext(), which ultimately frees sfp. After kref_put(), sg_release() continues to unlock open_rel_lock and may reference sfp or sdp. If sfp has already been freed, this results in a slab-use-after-free error. Move the kref_put(&sfp->f_ref, sg_remove_sfp) call after unlocking the open_rel_lock mutex. This ensures: - No references to sfp or sdp occur after the reference count is decremented. - Cleanup functions such as sg_remove_sfp() and sg_remove_sfp_usercontext() can safely execute without impacting the mutex handling in sg_release(). The fix has been tested and validated by syzbot. This patch closes the bug reported at the following syzkaller link and ensures proper sequencing of resource cleanup and mutex operations, eliminating the risk of use-after-free errors in sg_release(). Reported-by: syzbot+7efb5850a17ba6ce098b@syzkaller.appspotmail.com Closes: https://syzkaller.appspot.com/bug?extid=7efb5850a17ba6ce098b Tested-by: syzbot+7efb5850a17ba6ce098b@syzkaller.appspotmail.com Fixes: cc833acbee9d ("sg: O_EXCL and other lock handling") Signed-off-by: Suraj Sonawane Link: https://lore.kernel.org/r/20241120125944.88095-1-surajsonawane0215@gmail.com Reviewed-by: Bart Van Assche Signed-off-by: Martin K. Petersen

scsi: sg: Enable runtime power management

2024-11-07T01:52:28+00:00

In 2010, runtime power management support was implemented in the SCSI core. The description of patch "[SCSI] implement runtime Power Management" mentions that the sg driver is skipped but not why. This patch enables runtime power management even if an instance of the sg driver is held open. Enabling runtime PM for the sg driver is safe because all interactions of the sg driver with the SCSI device pass through the block layer (blk_execute_rq_nowait()) and the block layer already supports runtime PM. Cc: Alan Stern Cc: Douglas Gilbert Fixes: bc4f24014de5 ("[SCSI] implement runtime Power Management") Signed-off-by: Bart Van Assche Link: https://lore.kernel.org/r/20241030220310.1373569-1-bvanassche@acm.org Acked-by: Alan Stern Signed-off-by: Martin K. Petersen

[tree-wide] finally take no_llseek out

2024-09-27T15:18:43+00:00

no_llseek had been defined to NULL two years ago, in commit 868941b14441 ("fs: remove no_llseek") To quote that commit, At -rc1 we'll need do a mechanical removal of no_llseek - git grep -l -w no_llseek | grep -v porting.rst | while read i; do sed -i '/\/d' $i done would do it. Unfortunately, that hadn't been done. Linus, could you do that now, so that we could finally put that thing to rest? All instances are of the form .llseek = no_llseek, so it's obviously safe. Signed-off-by: Al Viro Signed-off-by: Linus Torvalds

scsi: sg: Avoid race in error handling & drop bogus warn

2024-04-04T22:22:12+00:00

Commit 27f58c04a8f4 ("scsi: sg: Avoid sg device teardown race") introduced an incorrect WARN_ON_ONCE() and missed a sequence where sg_device_destroy() was used after scsi_device_put(). sg_device_destroy() is accessing the parent scsi_device request_queue which will already be set to NULL when the preceding call to scsi_device_put() removed the last reference to the parent scsi_device. Drop the incorrect WARN_ON_ONCE() - allowing more than one concurrent access to the sg device - and make sure sg_device_destroy() is not used after scsi_device_put() in the error handling. Link: https://lore.kernel.org/all/5375B275-D137-4D5F-BE25-6AF8ACAE41EF@linux.ibm.com Fixes: 27f58c04a8f4 ("scsi: sg: Avoid sg device teardown race") Cc: stable@vger.kernel.org Signed-off-by: Alexander Wetzel Link: https://lore.kernel.org/r/20240401191038.18359-1-Alexander@wetzel-home.de Tested-by: Sachin Sant Reviewed-by: Bart Van Assche Signed-off-by: Martin K. Petersen

scsi: sg: Avoid sg device teardown race

2024-03-25T19:41:07+00:00

sg_remove_sfp_usercontext() must not use sg_device_destroy() after calling scsi_device_put(). sg_device_destroy() is accessing the parent scsi_device request_queue which will already be set to NULL when the preceding call to scsi_device_put() removed the last reference to the parent scsi_device. The resulting NULL pointer exception will then crash the kernel. Link: https://lore.kernel.org/r/20240305150509.23896-1-Alexander@wetzel-home.de Fixes: db59133e9279 ("scsi: sg: fix blktrace debugfs entries leakage") Cc: stable@vger.kernel.org Signed-off-by: Alexander Wetzel Link: https://lore.kernel.org/r/20240320213032.18221-1-Alexander@wetzel-home.de Reviewed-by: Bart Van Assche Signed-off-by: Martin K. Petersen