[PATCH printk v1 0/2] Fix ABBA deadlock for legacy printing with cpu_sync

[John Ogness]

Hi,

An RFC patch [0] was posted to fix an ABBA deadlock related to
nmi_backtrace() and dump_stack_lvl(). However, the underlying
problem was trying to perform legacy printing while holding the
printk_cpu_sync.

This series causes legacy printing to defer when holding the
printk_cpu_sync.

The first patch in the series is not necessary for the deadlock
fix, but was a redundancy I noticed while working on the series.

John Ogness

[0] https://lore.kernel.org/lkml/20240715232052.73eb7fb1@xxxxxxxxxxxxxxxxxxxx

John Ogness (2):
  printk: Remove redundant deferred check in vprintk()
  printk: Defer legacy printing when holding printk_cpu_sync

 kernel/printk/internal.h    |  6 ++++++
 kernel/printk/printk.c      |  5 +++++
 kernel/printk/printk_safe.c | 16 ++++++----------
 3 files changed, 17 insertions(+), 10 deletions(-)


base-commit: 4022ef25504db2fb79a2acac0afe9bac934f8dd6
-- 
2.39.5