'lots of little changes' merge done

Geoffrey Keating geoffk@apple.com
Wed Nov 8 01:29:00 GMT 2006


On 07/11/2006, at 3:00 PM, Daniel Berlin wrote:

> On 11/7/06, Geoffrey Keating <geoffk@apple.com> wrote:
>>
>> On 07/11/2006, at 9:38 AM, Daniel Berlin wrote:
>>
>> >> You need dwarf_utilities-25, I think.  I'm not sure where to get  
>> it
>> >> from for Tiger, I will inquire. It should be in the next Leopard  
>> seed
>> >> (if you have access to Leopard seeds).  In the meantime, you can
>> >> avoid the problem by copying /bin/true into your path as  
>> 'dsymutil';
>> >> the problem with that, or an autoconf test or whatever, is that if
>> >> you say 'gcc -g t.c -g -o t', debugging won't work.
>> >
>> > Just for the record, the workaround of symlinking /bin/true to
>> > dsymutil
>> > does not work.
>> >
>> > (We still fail bootstrap a litte later on).
>> > However, reverting the patch causes bootstap to succeed happily.
>>
>> What is the nature of the failure of the workaround?  The workaround
>> should completely disable the patch.
>
> It pulls dsymutil from /usr/bin anyway :(

Huh.  I wonder how it does that?

I guess you could replace /usr/bin/dsymutil with a shell script like:

#!/bin/sh
/usr/bin/dsymutil-old "$*"
exit 0

to suppress the failing error code.  That would probably be better  
than any of the previous suggestions because then you would still be  
able to debug programs without .o files.
-------------- next part --------------
A non-text attachment was scrubbed...
Name: smime.p7s
Type: application/pkcs7-signature
Size: 2462 bytes
Desc: not available
URL: <http://gcc.gnu.org/pipermail/gcc-patches/attachments/20061108/9df2f289/attachment.p7s>


More information about the Gcc-patches mailing list