Re: [for-linus][PATCH 5/5] recordmcount: Correct st_shndx handling

[Peter Zijlstra]

On Wed, Jun 09, 2021 at 08:33:49PM -0400, Steven Rostedt wrote:
> From: Peter Zijlstra <peterz@xxxxxxxxxxxxx>
> 
> One should only use st_shndx when >SHN_UNDEF and <SHN_LORESERVE. When
> SHN_XINDEX, then use .symtab_shndx. Otherwise use 0.
> 
> This handles the case: st_shndx >= SHN_LORESERVE && st_shndx != SHN_XINDEX.
> 
> Link: https://lkml.kernel.org/r/YL9HxEc/l0yrl5o8@xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx
> 
> Cc: stable@xxxxxxxxxxxxxxx
> Fixes: 4ef57b21d6fb4 ("recordmcount: support >64k sections")
> Reported-by: Mark-PK Tsai <mark-pk.tsai@xxxxxxxxxxxx>
> Tested-by: Mark-PK Tsai <mark-pk.tsai@xxxxxxxxxxxx>
> Acked-by: Ard Biesheuvel <ardb@xxxxxxxxxx>
> Signed-off-by: Peter Zijlstra (Intel) <peterz@xxxxxxxxxxxxx>
> Signed-off-by: Steven Rostedt (VMware) <rostedt@xxxxxxxxxxx>

This is apperently causing trouble for Stephen in -next. Please hold.