> This is the varargs code, and I currently solved it by using
> append_to_statement_list(), and then adding the resulting tree to the pre_p
> and post_p using gimplify_and_add(). Is it OK?
Take a look at mainstream ports (x86, rs6000, etc) and use them as a model.
--
Eric Botcazou
2009/5/16 Diego Novillo :
> On Fri, May 15, 2009 at 02:12, Eric Fisher wrote:
>
>> I think it's a mistake between a and b. It should be,
>>
>> "
>> variable @code{a} is completely modified with the contents of
>> variable @code{b}. Real definition are also known as @dfn{killing
>> definitions}.
> From: Jamie Prescott
> To: gcc@gcc.gnu.org
> Sent: Friday, May 15, 2009 3:28:18 PM
> Subject: Different reload behavior from 4.3.3 to 4.4
>
>
> In my VM, the X_REGS class is a generic 64bit regsiter class that can hold
> both 64bit DImode and 64bit DFmode.
> Such class does not allow direct c
> From: Jamie Prescott
> To: gcc@gcc.gnu.org
> Sent: Friday, May 15, 2009 4:16:30 PM
> Subject: Re: Different reload behavior from 4.3.3 to 4.4
>
>
> > From: Jamie Prescott
> > To: gcc@gcc.gnu.org
> > Sent: Friday, May 15, 2009 3:28:18 PM
> > Subject: Different reload behavior from 4.3.3 to 4.
> From: Jamie Prescott
> To: gcc@gcc.gnu.org
> Sent: Friday, May 15, 2009 3:28:18 PM
> Subject: Different reload behavior from 4.3.3 to 4.4
>
>
> In my VM, the X_REGS class is a generic 64bit regsiter class that can hold
> both 64bit DImode and 64bit DFmode.
> Such class does not allow direct c
In my VM, the X_REGS class is a generic 64bit regsiter class that can hold
both 64bit DImode and 64bit DFmode.
Such class does not allow direct constant loading, so in 4.3.3 I had:
enum reg_class
xxx_preferred_reload_class(rtx x, enum reg_class regclass)
{
enum machine_mode mode = GET_MOD
> From: Diego Novillo
> To: Jamie Prescott
> Cc: gcc@gcc.gnu.org
> Sent: Friday, May 15, 2009 2:40:15 PM
> Subject: Re: 4.4 API changes
>
> On Fri, May 15, 2009 at 17:23, Jamie Prescott wrote:
>
> > I managed to migrate my code to 4.4, but I've some code I cannot figure out
> > how to translat
> From: Diego Novillo
> To: Jamie Prescott
> Cc: gcc@gcc.gnu.org
> Sent: Friday, May 15, 2009 2:44:23 PM
> Subject: Re: Writing over [SP]
>
> On Thu, May 14, 2009 at 18:18, Jamie Prescott wrote:
> >
> > This is driving me crazy. Any memory op that tries to write over [SP] get's
> automaticall
On Fri, May 15, 2009 at 02:12, Eric Fisher wrote:
> I think it's a mistake between a and b. It should be,
>
> "
> variable @code{a} is completely modified with the contents of
> variable @code{b}. Real definition are also known as @dfn{killing
> definitions}. Similarly, the use of @code{b} read
On Thu, May 14, 2009 at 18:18, Jamie Prescott wrote:
>
> This is driving me crazy. Any memory op that tries to write over [SP] get's
> automatically nuked
> by the compiler.
> Any offset, positive or negative, from SP, has no problems.
> What did I do wrong this time?
You need to provide a *whol
On Fri, May 15, 2009 at 17:23, Jamie Prescott wrote:
> I managed to migrate my code to 4.4, but I've some code I cannot figure out
> how to translate.
> Before, I was building the TRUE and FALSE instruction list using
> gimplify_and_add(),
> and then a:
>
> build3(COND_EXPR, void_type_node, con
I managed to migrate my code to 4.4, but I've some code I cannot figure out
how to translate.
Before, I was building the TRUE and FALSE instruction list using
gimplify_and_add(),
and then a:
build3(COND_EXPR, void_type_node, cond, a_case, b_case);
Where 'a_case' and 'b_case' were tree's.
Now I
> To clarify, is it "make -l #" that fails or "make -j #" on Solaris?
"make -j"
> Parallel testing with -j# used to work fine, but admitedly its been a long
> while since I lost my solaris box... (I don't know if the load avg based
> mechanism for -l ever worked.) Is there a PR number?
That's
Michael Eager wrote:
> Ian Lance Taylor wrote:
>> Michael Eager writes:
>>
>>> Is there any documentation on the contents of .eh_frame
>>> and the augmentations used? IIRC, the data describes the
>>> try blocks and the catch handlers, but I'm looking for
>>> the gory details.
>>
>> I don't know o
On Thu, 14 May 2009, Eric Botcazou wrote:
> > The build went through without any error,
> > but most of the tests failed in "make check".
> > unexpected failures = 6472 and passed = 52.
>
> Try with "make -k check" and no -j, parallel testing is broken on Solaris.
> Eric Botcazou
To clarify, is i
Ian Lance Taylor wrote:
Michael Eager writes:
Is there any documentation on the contents of .eh_frame
and the augmentations used? IIRC, the data describes the
try blocks and the catch handlers, but I'm looking for
the gory details.
I don't know of any docs. Docs would be nice.
Here is the
Michael Eager writes:
> Is there any documentation on the contents of .eh_frame
> and the augmentations used? IIRC, the data describes the
> try blocks and the catch handlers, but I'm looking for
> the gory details.
I don't know of any docs. Docs would be nice.
Here is the comment from gcc/dw
Jamie Prescott writes:
> The target reorg is great, but it had two problems for me. One is that it was
> issued after the
> prologue/epilogue,
Oh yeah, you probably do want to introduce a pass before
prologue/epilogue generation but after register allocation. There isn't
any mechanism for that
Vladimir Makarov wrote:
People are complaining about GCC compilation speed and
of course we should work on its speedup. But GCC is not so bad, for
example SUN Studio compiler is almost 2 times slower than GCC.
Well, outside of these comparisons, I often completely flatten my
colleagues (at
josh.switni...@utoronto.ca writes:
> The build completes without error but when I try to compile a simple
> test program I get what I believe are newlib related errors.
>
>
> to eliminate error1:
> I added the newlib include directory manually to the include search path.
>
> "-I "
>
> and then I g
> What am I doing wrong?
You'll have to look at the logs and see whether there is a common pattern.
--
Eric Botcazou
- Original Message
> From: Ian Lance Taylor
> To: Jamie Prescott
> Cc: gcc@gcc.gnu.org
> Sent: Thursday, May 14, 2009 10:09:40 PM
> Subject: Re: Compact regsiter allocation
>
> Jamie Prescott writes:
>
> > If not, what is the best spot (in the normal GCC target hooks) to trigger
> >
Here's from the output of running "make -k check" (without "-j"),
notice that there's no # for libstdc++.
=== gcc Summary ===
# of expected passes18369
# of unexpected failures22530
# of expected failures 226
# of unresolved testcases 7716
# of
Quoting Ian Lance Taylor :
josh.switni...@utoronto.ca writes:
Josh, can you point Ian to where you put all your build logs?
here are all the build logs configure,make,install
http://code.google.com/p/tinyrtems/downloads/list
I don't see any "CC has changed" errors in those log files. I
josh.switni...@utoronto.ca writes:
>> Josh, can you point Ian to where you put all your build logs?
>
> here are all the build logs configure,make,install
>
> http://code.google.com/p/tinyrtems/downloads/list
I don't see any "CC has changed" errors in those log files. I don't see
any errors at a
Is there any documentation on the contents of .eh_frame
and the augmentations used? IIRC, the data describes the
try blocks and the catch handlers, but I'm looking for
the gory details.
--
Michael Eagerea...@eagercon.com
1960 Park Blvd., Palo Alto, CA 94306 650-325-8077
Quoting Joel Sherrill :
Ian Lance Taylor wrote:
Joel Sherrill writes:
Configuring in avr-elf/libgcc
configure: loading cache ./config.cache
configure: error: `CC' has changed since the previous run:
The change is that these options are new:
-nostdinc -B/home/cain/avr-tools/gcc/avr-elf/n
Quoting Bernd Schmidt :
2009-02-24 J"orn Rennecke
* loop-doloop.c (doloop_valid_p): Rename to:
(validize_doloop). Try to fix up loops with conditons for infinite
looping by enclosing them in an outer loop.
Changed caller.
(add_test): Add new parameter
Ian Lance Taylor wrote:
Joel Sherrill writes:
Configuring in avr-elf/libgcc
configure: loading cache ./config.cache
configure: error: `CC' has changed since the previous run:
The change is that these options are new:
-nostdinc -B/home/cain/avr-tools/gcc/avr-elf/newlib/ -isystem
/ho
Joern Rennecke wrote:
For ARCompact, doloop loops in particular were interesting to
optimize, which is why I have a patch to loop-doloop.c in
svn://gcc.gnu.org/svn/gcc/branches/arc-4_4-branch :
2009-02-24 J"orn Rennecke
* loop-doloop.c (doloop_valid_p): Rename to:
(validize
30 matches
Mail list logo
