Commit 35ebee7e7209 for kernel

commit 35ebee7e720944a66befb5899c72ce1e01dfa44e
Merge: 14df4eb7e7fa a0750fae73c5
Author: Linus Torvalds <torvalds@linux-foundation.org>
Date:   Fri Dec 12 22:04:18 2025 +1200

    Merge tag 'block-6.19-20251211' of git://git.kernel.org/pub/scm/linux/kernel/git/axboe/linux

    Pull block fixes from Jens Axboe:

     - Always initialize DMA state, fixing a potentially nasty issue on the
       block side

     - btrfs zoned write fix with cached zone reports

     - Fix corruption issues in bcache with chained bio's, and further make
       it clear that the chained IO handler is simply a marker, it's not
       code meant to be executed

     - Kill old code dealing with synchronous IO polling in the block layer,
       that has been dead for a long time. Only async polling is supported
       these days

     - Fix a lockdep issue in tag_set management, moving it to RCU

     - Fix an issue with ublks bio_vec iteration

     - Don't unconditionally enforce blocking issue of ublk control
       commands, allow some of them with non-blocking issue as they
       do not block

    * tag 'block-6.19-20251211' of git://git.kernel.org/pub/scm/linux/kernel/git/axboe/linux:
      blk-mq-dma: always initialize dma state
      blk-mq: delete task running check in blk_hctx_poll()
      block: fix cached zone reports on devices with native zone append
      block: Use RCU in blk_mq_[un]quiesce_tagset() instead of set->tag_list_lock
      ublk: don't mutate struct bio_vec in iteration
      block: prohibit calls to bio_chain_endio
      bcache: fix improper use of bi_end_io
      ublk: allow non-blocking ctrl cmds in IO_URING_F_NONBLOCK issue