On 02/14/2013 04:37 PM, Ralf Corsepius wrote:
>> At any rate, there’s a need for similar env. vars for cross-compilers. WDYT?
> I am having difficulties to imagine such a need.
Indeed: I've experienced that environment variables are fine to test something
out,
but open a can of worms when used
> -Original Message-
> From: Georg-Johann Lay [mailto:a...@gjlay.de]
> Sent: Wednesday, February 13, 2013 4:24 AM
> To: S, Pitchumani
> Cc: gcc@gcc.gnu.org; Denis Chertykov
> Subject: Re: Could not identify that register is clobbered already
>
> [Removing avr-gcc-list from CC because the
On 13-02-14 5:00 PM, Steven Bosscher wrote:
Hello Vlad,
Back in October, you added a patch to remove extra clobbers introduced by LRA:
http://gcc.gnu.org/ml/gcc-patches/2012-10/msg02531.html
This also disables some -frename-registers opportunities (a few 100 on
x86-64 for my collection of cc1-i
On 02/14/2013 02:41 PM, Michael Veksler wrote:
...
It sounds that your model requires many 0,1 variables. If it is so,
then maybe SAT
is a better choice than ILP.
Yes, it is binary ILP.
I don't see any progress, ILP solver may be ten times faster but they
still have exponential complexit
Status
==
GCC trunk remains in release branch mode, with only regression fixes
and documentation changes allowed.
Stabilisation for 4.8 release continues, with many regressions having
been fixed over the past month, although there are still seven P1 bugs
open for 4.8 and many other regression
Hello Vlad,
Back in October, you added a patch to remove extra clobbers introduced by LRA:
http://gcc.gnu.org/ml/gcc-patches/2012-10/msg02531.html
This also disables some -frename-registers opportunities (a few 100 on
x86-64 for my collection of cc1-i files).
Do you think it's be possible (and w
On 02/14/2013 06:31 PM, Vladimir Makarov wrote:
On 02/14/2013 03:46 AM, Michael Veksler wrote:
[snip]
I am reading this thread and getting more and more puzzled. The RA
stuff is very complicated,
having many constraints and many dependencies with other passes.
Taking this into
account, it seem
On 02/14/2013 03:46 AM, Michael Veksler wrote:
On 02/14/2013 03:28 AM, Vladimir Makarov wrote:
On 13-02-13 6:36 PM, Michael Eager wrote:
[snip]
I thought about register pressure causing this, but I think that
should cause
spilling of one of the registers which were not used in this long
se
On 02/13/2013 03:49 PM, Ludovic Courtès wrote:
Richard Biener skribis:
On Wed, Feb 13, 2013 at 12:13 PM, Richard Biener
wrote:
On Tue, Feb 12, 2013 at 9:41 PM, Ludovic Courtès wrote:
Joel Sherrill skribis:
But it still doesn't address the situation where you have multiple
cross compiler
> Isn't the REG_SETJMP note sufficient for this purpose?
Yeah, missed that. Sorry for flood. Thanks a lot!
> Could anybody pls advise, if I can detect that given RTL `call` is
> actually a setjmp ()?
>
> I see no references in dump...
> (call_insn 6 5 7 (set (reg:SI 0 ax)
> (call (mem:QI (symbol_ref:DI ("_setjmp") [flags 0x41]
> ) [0 _setjmp S1 A8])
> (const_int 0 [0]))) 4.c:17 -1
>
On Thu, Feb 14, 2013 at 9:35 AM, Kirill Yukhin wrote:
> Hi,
> Could anybody pls advise, if I can detect that given RTL `call` is
> actually a setjmp ()?
Get at the function-decl and use special_function_p (decl, 0) &
ECF_RETURNS_TWICE
(that will conservatively detect all possible setjmp calls but
On 02/14/2013 03:28 AM, Vladimir Makarov wrote:
On 13-02-13 6:36 PM, Michael Eager wrote:
[snip]
I thought about register pressure causing this, but I think that
should cause
spilling of one of the registers which were not used in this long
sequence,
rather than causing a large number of a
Hi,
Could anybody pls advise, if I can detect that given RTL `call` is
actually a setjmp ()?
I see no references in dump...
(call_insn 6 5 7 (set (reg:SI 0 ax)
(call (mem:QI (symbol_ref:DI ("_setjmp") [flags 0x41]
) [0 _setjmp S1 A8])
(const_int 0 [0]))) 4.c:17 -1
(expr_li
14 matches
Mail list logo
