pointer comparison breaks Solaris' qsort

Alexandre Oliva aoliva@redhat.com
Wed Jun 9 13:37:00 GMT 2004


On Jun  9, 2004, Alan Modra <amodra@bigpond.net.au> wrote:

> On Wed, Jun 09, 2004 at 12:35:03PM +0100, Dave Korn wrote:
>> > -----Original Message-----
>> > From: binutils-owner On Behalf Of Alexandre Oliva
>> > Sent: 09 June 2004 12:29
>> 
>> > Maybe if you can point me at the other chunk of code you have in mind,
>> > I could volunteer to fix this other bug, but conditioning the
>> > acceptance of a correct patch that fixes a bug on fixing some other
>> > bug is not reasonable IMHO.
>> 
>> .... unless consistency between the two pieces of code is a vital
>> requirement, n'est-ce pas?

> Exactly.  Alex, the other chunk of code is in the block immediately
> after the qsort that uses elf_sort_symbol.

Aah!  I was looking for another qsort compare function.  Sorry.
Somehow I missed the `binary search' reference in your e-mail.  Please
accept my apologies for the tone of my previous e-mail.  It was not
called for.

> If you sort the array differently (your change to elf_sort_symbol),
> you'd better also do the binary search in the same way.

Certainly.

Here's the patch I'm testing atm.  I'll check it in if testing
succeeds.

Thanks for pointing out my mistake.

-------------- next part --------------
A non-text attachment was scrubbed...
Name: bfd-elf-sort-symbol-solaris-crash.patch
Type: text/x-patch
Size: 1028 bytes
Desc: not available
URL: <https://sourceware.org/pipermail/binutils/attachments/20040609/d32b182c/attachment.bin>
-------------- next part --------------

-- 
Alexandre Oliva             http://www.ic.unicamp.br/~oliva/
Red Hat Compiler Engineer   aoliva@{redhat.com, gcc.gnu.org}
Free Software Evangelist  oliva@{lsd.ic.unicamp.br, gnu.org}


More information about the Binutils mailing list