RFC jump label v4
Get series with b4
b4 am [email protected]Versions newest first
- v1 RFC Latest8 patches 26 repliesIn review
Patches in v1 No applies recorded yet
| # | Subject and files | Size | Replies | Applied |
|---|---|---|---|---|
| 1/8 | jump label v4 - kprobes/x86: Cleanup RELATIVEJUMP_INSTRUCTION to RELATIVEJUMP_OPCODEarch/ arch/ | +2 | 0 replies | |
| 2/8 | jump label v4 - x86: Introduce generic jump patching without stop_machinearch/ arch/ arch/ +1 more | +134 | 26 replies | |
| 3/8 | jump label v4 - move opcode definitionsarch/ arch/ | +5 | 0 replies | |
| 4/8 | jump label v4 - notifier atomic call chain notracekernel/ | +3 | 0 replies | |
| 5/8 | jump label v4 - base patchinclude/ include/ kernel/ +1 more | +232 | 0 replies | |
| 6/8 | jump label v4 - x86 supportarch/ arch/ arch/ | +85 | 0 replies | |
| 7/8 | jump label v4 - tracepoint supportinclude/ kernel/ | +27 | 0 replies | |
| 8/8 | jump label v4 - add module supportarch/ include/ include/ +2 more | +150 | 0 replies |
Reviews trailers per patch
No Reviewed-by, Acked-by or Tested-by trailers for v1 yet.
Cover letter [RFC PATCH 0/8] jump label v4
hi,
For background, see previous posting: http://marc.info/?l=linux-kernel&m=125858436505941&w=2
Refresh of the jump labeling patches. The largest update revolves around how the jump labels are organized. I've added a hash table, using the name of the tracepoints as the 'key'. (I might want to change this to a pointer, for other users, but the string key works well for tracepoints). I've also pre-sorted the jump tables by name. Thus, when a jump label is enabled/disabled we look up its hash table entry and then walk the list of associated jump labels. The implementation also associates module jump sections with the appropriate keys as they are inserted and removed.
The first 2 patches of the series are a repost of Masami's text_poke_fixup() function, which allows for efficient instruction patching. I believe this is still an ongoing discussion, on the best safest approach here. Masami's text_poke_fixup() has been working well for me. But I could certainly substitute a 'stop_machine()' version or something similar if there is enough interest, until we have a more efficient solution.
patches are against the latest -tip tree.
thanks,
-Jason
Masami Hiramatsu (2):
x86: Introduce generic jump patching without stop_machine kprobes/x86: Cleanup RELATIVEJUMP_INSTRUCTION to RELATIVEJUMP_OPCODE
Mathieu Desnoyers (1):
notifier atomic call chain notrace
Jason Baron(5):
move opcode defs from asm/kprobes.h to asm/alternative.h jump label base
x86: jump label support
jump label tracepoint support
jump label module support
arch/x86/include/asm/alternative.h | 17 ++ arch/x86/include/asm/jump_label.h | 32 ++++ arch/x86/include/asm/kprobes.h | 3 - arch/x86/kernel/Makefile | 2 +- arch/x86/kernel/alternative.c | 120 ++++++++++++++ arch/x86/kernel/jump_label.c | 52 ++++++ arch/x86/kernel/kprobes.c | 2 +- arch/x86/kernel/ptrace.c | 1 + include/asm-generic/vmlinux.lds.h | 10 +- include/linux/jump_label.h | 58 +++++++ include/linux/module.h | 5 +- include/linux/tracepoint.h | 34 +++-- kernel/Makefile | 2 +- kernel/jump_label.c | 301 ++++++++++++++++++++++++++++++++++++ kernel/kprobes.c | 2 +- kernel/module.c | 7 + kernel/notifier.c | 6 +- kernel/tracepoint.c | 8 + 18 files changed, 635 insertions(+), 27 deletions(-) create mode 100644 arch/x86/include/asm/jump_label.h create mode 100644 arch/x86/kernel/jump_label.c create mode 100644 include/linux/jump_label.h create mode 100644 kernel/jump_label.c