[llvm-testresults] buildbot failure in smooshlab on llvm-gcc-i386-darwin9
Dale Johannesen
dalej at apple.com
Wed Nov 17 19:39:46 PST 2010
If Devang told you that he is right. Ignore me.
On Nov 17, 2010, at 7:37 PM, Rafael EspĂndola wrote:
> 2010/11/17 Dale Johannesen <dalej at apple.com>:
>> Probably not, that was the shipping assembler on darwin 9 (Leopard).
>> I think we want to continue working in that environment.
>
> So, it should be possible to print the old DW_LNE_set_address based
> line table in the finish method of the asm streamer, but I was told by
> daniel to not worry about Darwin and Devang told me:
>
> Here we are assuming that all supported platforms' assembler supports
> .loc/.file, which is ok.
>
> So, do I have to implement the DW_LNE_set_address fallback in MC? Is
> it sufficient to just not print debug lines in old system?
>
> Note that this doesn't affect clang as clang uses the integrated
> assembler.
>
> Cheers,
> Rafael
More information about the llvm-testresults
mailing list