linux-toradex.git/drivers/scsi/scsi_scan.c, branch v6.16

scsi: core: Fix missing lock protection

2025-03-03T02:06:02+00:00

async_scan_lock is designed to protect the scanning_hosts list, but there
is no protection here.

Signed-off-by: Chaohai Chen 
Link: https://lore.kernel.org/r/20250221030755.219277-1-wdhh66@163.com
Reviewed-by: Bart Van Assche 
Signed-off-by: Martin K. Petersen

scsi: core: Use GFP_NOIO to avoid circular locking dependency

2025-02-03T22:08:13+00:00

Filesystems can write to disk from page reclaim with __GFP_FS
set. Marc found a case where scsi_realloc_sdev_budget_map() ends up in
page reclaim with GFP_KERNEL, where it could try to take filesystem
locks again, leading to a deadlock.

WARNING: possible circular locking dependency detected
6.13.0 #1 Not tainted
------------------------------------------------------
kswapd0/70 is trying to acquire lock:
ffff8881025d5d78 (&q->q_usage_counter(io)){++++}-{0:0}, at: blk_mq_submit_bio+0x461/0x6e0

but task is already holding lock:
ffffffff81ef5f40 (fs_reclaim){+.+.}-{0:0}, at: balance_pgdat+0x9f/0x760

The full lockdep splat can be found in Marc's report:

https://lkml.org/lkml/2025/1/24/1101

Avoid the potential deadlock by doing the allocation with GFP_NOIO, which
prevents both filesystem and block layer recursion.

Reported-by: Marc Aurèle La France 
Signed-off-by: Rik van Riel 
Link: https://lore.kernel.org/r/20250129104525.0ae8421e@fangorn
Reviewed-by: Christoph Hellwig 
Signed-off-by: Martin K. Petersen

Merge tag 'block-6.14-20250131' of git://git.kernel.dk/linux

2025-01-31T19:49:30+00:00

Pull more block updates from Jens Axboe:

 - MD pull request via Song:
      - Fix a md-cluster regression introduced

 - More sysfs race fixes

 - Mark anything inside queue freezing as not being able to do IO for
   memory allocations

 - Fix for a regression introduced in loop in this merge window

 - Fix for a regression in queue mapping setups introduced in this merge
   window

 - Fix for the block dio fops attempting an iov_iter revert upton
   getting -EIOCBQUEUED on the read side. This one is going to stable as
   well

* tag 'block-6.14-20250131' of git://git.kernel.dk/linux:
  block: force noio scope in blk_mq_freeze_queue
  block: fix nr_hw_queue update racing with disk addition/removal
  block: get rid of request queue ->sysfs_dir_lock
  loop: don't clear LO_FLAGS_PARTSCAN on LOOP_SET_STATUS{,64}
  md/md-bitmap: Synchronize bitmap_get_stats() with bitmap lifetime
  blk-mq: create correct map for fallback case
  block: don't revert iter for -EIOCBQUEUED

block: force noio scope in blk_mq_freeze_queue

2025-01-31T14:20:08+00:00

When block drivers or the core block code perform allocations with a
frozen queue, this could try to recurse into the block device to
reclaim memory and deadlock.  Thus all allocations done by a process
that froze a queue need to be done without __GFP_IO and __GFP_FS.
Instead of tying to track all of them down, force a noio scope as
part of freezing the queue.

Note that nvme is a bit of a mess here due to the non-owner freezes,
and they will be addressed separately.

Signed-off-by: Christoph Hellwig 
Link: https://lore.kernel.org/r/20250131120352.1315351-2-hch@lst.de
Signed-off-by: Jens Axboe

scsi: scsi_scan: Add kernel-doc for exported function

2025-01-10T20:15:16+00:00

Add kernel-doc for scsi_add_device() since it is exported.  This allows it
to be part of the SCSI driver-api docbook.

Signed-off-by: Randy Dunlap 
Link: https://lore.kernel.org/r/20241212205217.597844-5-rdunlap@infradead.org
CC: James E.J. Bottomley 
CC: Martin K. Petersen 
Signed-off-by: Martin K. Petersen

scsi: core: Remove the .slave_configure() method

2024-12-04T20:34:28+00:00

Now that all SCSI drivers have been converted from .slave_configure() to
.sdev_configure(), remove support for .slave_configure() from the SCSI
core.

Reviewed-by: Damien Le Moal 
Signed-off-by: Bart Van Assche 
Link: https://lore.kernel.org/r/20241022180839.2712439-5-bvanassche@acm.org
Signed-off-by: Martin K. Petersen

scsi: Convert SCSI drivers to .sdev_configure()

2024-12-04T20:34:28+00:00

The only difference between the .sdev_configure() and .slave_configure()
methods is that the former accepts an additional 'limits' argument.
Convert all SCSI drivers that define a .slave_configure() method to
.sdev_configure(). This patch prepares for removing the
.slave_configure() method. No functionality has been changed.

Acked-by: Geoff Levand  # for ps3rom
Acked-by: Khalid Aziz  # for the BusLogic driver
Reviewed-by: Damien Le Moal 
Signed-off-by: Bart Van Assche 
Link: https://lore.kernel.org/r/20241022180839.2712439-4-bvanassche@acm.org
Signed-off-by: Martin K. Petersen

scsi: Rename .device_configure() into .sdev_configure()

2024-12-04T20:34:28+00:00

Improve naming consistency with the .sdev_prep() and .sdev_destroy()
methods by renaming .device_configure() into .sdev_configure().

Cc: Christoph Hellwig 
Acked-by: Damien Le Moal 
Signed-off-by: Bart Van Assche 
Link: https://lore.kernel.org/r/20241022180839.2712439-3-bvanassche@acm.org
Signed-off-by: Martin K. Petersen

scsi: Rename .slave_alloc() and .slave_destroy()

2024-12-04T20:34:28+00:00

Rename .slave_alloc() into .sdev_init() and .slave_destroy() into
.sdev_destroy(). The new names make it clear that these are actions on
SCSI devices. Make this change in the SCSI core, SCSI drivers and also
in the ATA drivers. No functionality has been changed.

This patch has been created as follows:
* Change the text "slave_alloc" into "sdev_init" in all source files
  except those in drivers/net/ and Documentation/.
* Change the text "slave_destroy" into "sdev_destroy" in all source
  files except those in drivers/net/ and Documentation/.
* Rename lpfc_no_slave() into lpfc_no_sdev().
* Manually adjust whitespace where necessary to restore vertical
  alignment (dc395x driver and include/linux/libata.h).

Acked-by: Damien Le Moal 
Signed-off-by: Bart Van Assche 
Link: https://lore.kernel.org/r/20241022180839.2712439-2-bvanassche@acm.org
Signed-off-by: Martin K. Petersen

move asm/unaligned.h to linux/unaligned.h

2024-10-02T21:23:23+00:00

asm/unaligned.h is always an include of asm-generic/unaligned.h;
might as well move that thing to linux/unaligned.h and include
that - there's nothing arch-specific in that header.

auto-generated by the following:

for i in `git grep -l -w asm/unaligned.h`; do
	sed -i -e "s/asm\/unaligned.h/linux\/unaligned.h/" $i
done
for i in `git grep -l -w asm-generic/unaligned.h`; do
	sed -i -e "s/asm-generic\/unaligned.h/linux\/unaligned.h/" $i
done
git mv include/asm-generic/unaligned.h include/linux/unaligned.h
git mv tools/include/asm-generic/unaligned.h tools/include/linux/unaligned.h
sed -i -e "/unaligned.h/d" include/asm-generic/Kbuild
sed -i -e "s/__ASM_GENERIC/__LINUX/" include/linux/unaligned.h tools/include/linux/unaligned.h