Re: [Patches] [patch] Fix for debug/tst-backtrace{5, 6} failure of 32-bit libc on 64-bit host

["Joseph S. Myers" <joseph@xxxxxxxxxxxxxxxx>]

On Mon, 20 Feb 2012, Paul Pluzhnikov wrote:

> The man page and backtrace itself could of course be changed, but (I expect)
> doing so will break many applications that are already doing -1 on return
> addresses.
> 
> An alternate fix could be to re-try symbol @RA-1 if symbol @RA is empty
> in tst-backtrace6.c.

I'd have thought backtrace_symbols (and backtrace_symbols_fd) would be a 
better place for that logic.  Though ideally to get backtraces right in 
all cases you'd have a version of backtrace that calls _Unwind_GetIPInfo 
(if available) instead of _Unwind_GetIP so the correct adjustment can be 
made.  In which case you'd need separate versions of backtrace_symbols 
etc. as well for the adjusted backtraces, if backtrace_symbols and 
backtrace_symbols_fd do adjustment themselves.  Maybe raise on libc-alpha 
the general question of how to get reliable backtraces - which need 
addresses within the right function - with the backtrace function or a new 
variant thereof?

-- 
Joseph S. Myers
joseph@xxxxxxxxxxxxxxxx
_______________________________________________
Patches mailing list
Patches@xxxxxxxxxx
http://eglibc.org/cgi-bin/mailman/listinfo/patches