As pointed out in the review [1] of commit 39d0be87438a ("serial:
kgdb_nmi: Remove unused knock code"), it's not just the knock code
that's unused but a whole chunk of NMI-related code in KGDB.
Look back at the original series [2] and revert a few of the patches
that managed to land and had no other purpose. They've been in the
kernel for over 12 years and if nobody has used them during that time
then they won't get used.
I've validated things still compile just fine for me and I've tested
"allmodconfig" for a few different architectures. Hopefully I didn't
miss anything, but based on "grep" this is all just dead code.
I guess I'd expect this to land through Greg KH's tree, ideally Acked
by Daniel Thompson.
[1] https://lore.kernel.org/r/CAD=FV=VH5YWfTggGNt2KLoSPKqVTmWArEgccok3uFQ5BCA7cNw@xxxxxxxxxxxxxx
[2] https://lore.kernel.org/lkml/1348522080-32629-9-git-send-email-anton.vorontsov@xxxxxxxxxx/
Douglas Anderson (3):
Revert "tty/serial: Add kgdb_nmi driver"
Revert "kdb: Implement disable_nmi command"
Revert "kernel/debug: Mask KGDB NMI upon entry"
drivers/tty/serial/Kconfig | 19 ---
drivers/tty/serial/Makefile | 1 -
drivers/tty/serial/kgdb_nmi.c | 280 ----------------------------------
drivers/tty/serial/kgdboc.c | 8 -
include/linux/kgdb.h | 11 --
kernel/debug/debug_core.c | 14 +-
kernel/debug/kdb/kdb_main.c | 37 -----
7 files changed, 3 insertions(+), 367 deletions(-)
delete mode 100644 drivers/tty/serial/kgdb_nmi.c
--
2.48.1.262.g85cc9f2d1e-goog
