On 11/08/14 09:16, Iain Sandoe wrote:
Hi Jeff,
On 7 Nov 2014, at 20:28, Jeff Law wrote:
On 11/06/14 06:01, Evgeny Stupachenko wrote:
Now I see that equiv reload could be special for PIC register.
Let's apply more conservative patch.
Darwin bootstrap passed with the patch applied on r216304 (along
with already committed to trunk patches from PR63618 and
PR63620).
2014-11-06 Evgeny Stupachenko <evstu...@gmail.com>
PR target/63534 * config/i386/i386.c (builtin_setjmp_receiver):
Use pic_offset_table_rtx for PIC register.
(nonlocal_goto_receiver): Delete.
OK for the trunk. One more Darwin bug gets squashed :-)
FOAD, are you referring to this bug (63534) - or is there an
implication that the nonlocal_goto_reciever was buggy (*before* the
trunk changes) - I ask because the same implementation is present on
the open branches, and if it's broken would like to sort that out.
cheers Iain
No need to do anything on the branches as the issue with the
{setjmp,nonlocal_goto}_receiver is specific to the change to expose the
PIC register to register allocation.
jeff
