kallsyms should prefer non weak symbols
commitf2df3f65d0b4337cfb5b19eab3ee28b177427c49
authorPaulo Marques <pmarques@grupopie.com>
Wed, 6 Feb 2008 09:37:33 +0000 (6 01:37 -0800)
committerLinus Torvalds <torvalds@woody.linux-foundation.org>
Wed, 6 Feb 2008 18:41:07 +0000 (6 10:41 -0800)
tree86d8aae1e9ef8c57721c886f67a915ba1897dc69
parenta6752f3f538e9dc0d0e7fdb2080532554a5eb395
kallsyms should prefer non weak symbols

When resolving symbol names from addresses with aliased symbol names,
kallsyms_lookup always returns the first symbol, even if it is a weak
symbol.

This patch changes this by sorting the symbols with the weak symbols last
before feeding them to the kernel.  This way the kernel runtime isn't
changed at all, only the kallsyms build system is changed.

Another side effect is that the symbols get sorted by address, too.  So,
even if future binutils version have some bug in "nm" that makes it fail to
correctly sort symbols by address, the kernel won't be affected by this.

Mathieu says:

  I created a module in LTTng that uses kallsyms to get the symbol
  corresponding to a specific system call address.  Unfortunately, all the
  unimplemented syscalls were all referring to the (same) weak symbol
  identifying an unrelated system call rather that sys_ni (or whatever
  non-weak symbol would be expected).  Kallsyms was dumbly returning the first
  symbol that matched.

  This patch makes sure kallsyms returns the non-weak symbol when there is
  one, which seems to be the expected result.

[akpm@linux-foundation.org: coding-style fixes]
Signed-off-by: Mathieu Desnoyers <mathieu.desnoyers@polymtl.ca>
Looks-great-to: Rusty Russell <rusty@rustcorp.com.au>
Cc: Sam Ravnborg <sam@ravnborg.org>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
scripts/kallsyms.c