This is the mail archive of the gcc-patches@gcc.gnu.org mailing list for the GCC project.


Index Nav: [Date Index] [Subject Index] [Author Index] [Thread Index]
Message Nav: [Date Prev] [Date Next] [Thread Prev] [Thread Next]
Other format: [Raw text]

Re: [RFA:] Split REG_LABEL into REG_LABEL_TARGET and REG_LABEL_OPERAND


> :ADDPATCH middle-end:

rtl-optimization, otherwise I cannot approve it. :-)

> Testing the instrumented code on native x86_64-unknown-linux-gnu
> and crosses to mips-elf, sh64-elf(!), cris-elf showed no
> regressions.  But for sh-elf, it trigged in sched1 when building
> newlib; a label load was feeding a casesi insn.  The combination
> seemed valid, but I was a bit surprised to see it not trig for
> sh64.

Thanks for conducting the experiment.

> 	* sched-rgn.c (is_cfg_nonregular): Don't return 1 for a
> 	single_set insn only feeding a label to a jump through a
> 	register that dies there.

OK.


:REVIEWMAIL:

-- 
Eric Botcazou


Index Nav: [Date Index] [Subject Index] [Author Index] [Thread Index]
Message Nav: [Date Prev] [Date Next] [Thread Prev] [Thread Next]