as the Intel
> compiler) as is reasonable.
Yes, I don't think we should require GCC to build GCC, this would be a step
backwards in my opinion. I can experiment with the Sun Studio compiler.
--
Eric Botcazou
rc/gcc/ada/gcc-interface/decl.c:7992:33: warning:
identifier 'new' conflicts with C++ keyword
/home/eric/gnat/gnat-head/src/gcc/ada/gcc-interface/decl.c:8008:9: warning:
identifier 'new' conflicts with C++ keyword
but they don't stop the build because -Werror is not passed. That needs to be
fixed first.
--
Eric Botcazou
> This was the only va_arg usage, may be we should apply it on trunk too
> as the patched version is supposed to work for both C and C++.
Yes, but I'm testing a patch for trunk with more changes.
--
Eric Botcazou
had
been reported for weeks in PRs and the problematic patch clearly identified.
--
Eric Botcazou
> and read much more detailed release notes here:
> http://llvm.org/releases/2.7/docs/ReleaseNotes.html
The correct writing of GCC is "GCC", all capitalized. In particular "gcc-4.5"
should be written GCC 4.5 (like LLVM 2.7).
--
Eric Botcazou
> I agree that this really ought to be documented.
There is an old PR about this: http://gcc.gnu.org/PR16519
--
Eric Botcazou
lt in it being lost and forgotten. In order to report
an issue, please open a ticket with bugzilla. In order to submit a patch,
please use gcc-patc...@gcc.gnu.org. In both cases, follow the guidelines
written down in the aforementioned documentation. Thanks in advance.
--
Eric Botcazou
? Thanks in advance.
--
Eric Botcazou
for this would be lto_symtab_merge_cgraph_nodes () which already
> is after type merging and still has all candidates available.
Is suppressing the diagnostic sufficient? Will the two types be merged after
the symbols are unified, or does that not matter at all?
--
Eric Botcazou
14.56 11.29%
garbage_collection 0.28 0.07%
total 383.21100.00%
--
Eric Botcazou
nal5.26
symout 8.94
14.56 3.80%
garbage_collection 0.28 0.07%
total 383.21100.00%
--
Eric Botcazou
(3) the result of the load is unused when the memory
location is uninitialized. So, from an external viewpoint, the generated
code behaves as if there were no "problematic" loads and looks therefore OK.
--
Eric Botcazou
You don't need several personality routines to compile an all-Ada
program. This can presumably happen for an all-C++ program as well, but is
masked if you have recent enough binutils (2.20 and above).
--
Eric Botcazou
uld by name).
Can't we simply record the first personality routine encountered when reading
in the GIMPLE IL and assign it to all FDEs that don't have one? That would
at least solve the problem in the homogeneous case.
--
Eric Botcazou
me and deduce something from its presence in the scope.
This was done this way to prevent debug stuff from altering code generation
and avoid wasting space. Yes, the whole design is questionable, give me a
few days to investigate whether it can be modified.
--
Eric Botcazou
d by the configure machinery at each stage into the
appropriate values, which vary with each stage.
--
Eric Botcazou
ike we currently require only an ISO C90 compiler.
--
Eric Botcazou
> C90 does not have long long either, yet we use it as required (for HWI).
No, we use it when present but we don't require it, see hwint.h.
--
Eric Botcazou
+, that's orthogonal.
--
Eric Botcazou
o fancy C++ style, but Mark
was clear about that. Maintainers should be allowed to reject conversions
that buy too little or nothing.
--
Eric Botcazou
> I'd like us to stick with C comments only. I defintely do not like
> a mix of both styles and I can't see an advantage of C++ comments.
It indeed seems to fall into the buy-us-nothing category.
--
Eric Botcazou
was this committed" messages though, that's
rather annoying. The gcc-cvs archives are http://gcc.gnu.org/ml/gcc-cvs/
--
Eric Botcazou
s to
> effectively close threads, and then you'll see very few messages from me.
Please no, that's just a sheer waste of time/bandwidth/storage.
--
Eric Botcazou
e ChangeLog (and optionally the URL).
--
Eric Botcazou
> Are you volunteering to write that small script?
If nothing better comes out, why not, but resurrecting the Patch Tracker seems
to be a more appealing idea.
--
Eric Botcazou
he latter to dynamically affect the former. That's
certainly doable but not immediate.
--
Eric Botcazou
gnificantly. The attached patchlet alone saves 3% in code size
at -Os on a 50 MB executable and yields a 5% speedup at -O2 on another code.
* ipa-inline.c (likely_eliminated_by_inlining_p): Really consider that
loads from parameters passed by reference are free after inlining.
;& (is_gimple_reg (lhs)
! || !is_gimple_reg_type (TREE_TYPE (lhs
lhs_free = true;
> Which you could have simplified to just say
>
> if (lhs_free || rhs_free)
> return true;
>
> and drop the code you are changing.
I don't think so, compare your version and mine for scalar stores/loads
from/to parameters or return values.
--
Eric Botcazou
s that all scalar (thus gimple_reg_typed)
> loads/stores to/from parameters or the result are free.
Precisely not, they aren't free, otherwise they wouldn't exist in the first
place. Scalar loads/stores are never free, aggregate loads/stores may be
free if they are created only to pass the object around.
--
Eric Botcazou
> Any ideas what might be wrong here?
Probably PR libgcj/44415.
--
Eric Botcazou
ended versions (GMP 4.3.2, MPFR 2.4.2, MPC 0.8.1) also work fine if
compiled by the latest GCC 4.3.x compilers (or compilers of later series).
--
Eric Botcazou
==
>
> # of expected passes20118
> # of unexpected failures29
> # of unexpected successes 4
> # of expected failures 158
> # of unsupported tests 136
I meant "make check" for GMP and MPFR, not for GCC.
--
Eric Botcazou
h. You should probably wait a few days to
let people comment and/or fix, and then file PRs. 1 per file seems to be the
right granularity.
* rtlanal.c (nonzero_bits1): Use unsigned HOST_WIDE_INT in all mask
computations. Fix formatting is
y changes in a way that breaks the gcc build we
> notice. But I don't think we have any mechanism to do that at
> present.
Yes, I think we should have recommended versions for the 3rd party libraries.
--
Eric Botcazou
tart to post their patches on the gcc-patches@ list. I'll sponsor them
for write access at that point.
--
Eric Botcazou
erything is documented on http://gcc.gnu.org/contribute.html
Ian, would you mind sending them the appropriate form? Thanks in advance.
--
Eric Botcazou
> The performance testing didn't show performance regressions.
> 4.4.5 RC1 is currently blocked by PR45234, either it needs to be fixed
> soon, or reverted.
I'd like to backport the fix for middle-end/44763 before the RC. May I?
--
Eric Botcazou
> Sure.
Thanks, done after bootstrap/regtesting on x86_64-suse-linux.
--
Eric Botcazou
intainers to respond might take forever), and 2. make a broad
> announcement on gcc-patches with a list of targets modified, and give their
> maintainers a week to speak up if they have remarks. Is that enough? Is
> that overkill?
I think you need a global reviewer for broad changes like this.
--
Eric Botcazou
f the compiler itself.
The compiler driver program runs one or more of the subprograms
`cpp', `cc1', `as' and `ld'. It tries PREFIX as a prefix for each
program it tries to run, both with and without `MACHINE/VERSION/'
(*note Target Options::).
--
Eric Botcazou
case.
This PR is a pathological case that doesn't come from real life and fixing it
properly is hard, so I wouldn't bother about it.
--
Eric Botcazou
run a script to "link"
all the VCG files for a program and you can (try and) compute the worst case
stack usage for example. This works reasonably well for embedded stuff.
I'll try and submit something before stage 1 ends if people are interested.
--
Eric Botcazou
it is proper if we remove this testsuite from the
> GCC testsuites?
Marking it as XFAIL is the usual way of dealing with such issues.
--
Eric Botcazou
ard-to-debug problems exposed by Ada.
--
Eric Botcazou
true in 32-bit mode is partly used to implement
the psABI; you cannot simply toggle the return value, further adjustments are
needed to compensate. But this should ultimately work, as in 64-bit mode.
Are there still any good reasons to keep the old setting at this point?
--
Eric Botcazou
tures in the
same LTO compilation, but also updating the current architecture for later
write operations. This is modelled on the linker.
--
Eric Botcazou
use.
There should be nothing special to do, as the control flow transfer should be
represented in the RTL by means of a control_flow_insn_p insn. Perhaps the
related change made by Bernd in rev 162301 should be extended to the JUMP_P
block a few lines below?
--
Eric Botcazou
> JUMP_INSNs already invalidate the register use information. The problem
> is that CALL_INSNs that can throw don't.
Sure, that's precisely what I was suggesting to change, like in rev 162301.
--
Eric Botcazou
TERN (insn)) != RETURN)
{
/* Non-spill registers might be used at the call destination in
some unknown fashion, so we have to mark the unknown use. */
--
Eric Botcazou
ompiler to select the right variant. --with-cpu isn't appropriate for this.
It turns out that we have ported GNAT (the GNU Ada compiler) to embedded LEON
targets at AdaCore so we also have some material, mostly orthogonal to yours.
Give me a few days to port your 4.4 patches and dig out ours, and I'll post a
combined patch to serve as a basis for further discussions.
--
Eric Botcazou
; Maybe there is a simple way to achieve both multilib and singlelib?
You can pass --disable-multilib at configure time.
--
Eric Botcazou
n is also straightforward: LEON is an implementation
of the SPARC-V8 architecture so --with-cpu=v8 and --with-float=hard.
> Also, it might happen that someone doesn't want one multilib dimension, but
> they want to keep another one.
Indeed, being able to partially disable multilibs woul
ilib stuff is for the
default target sparc-leon-elf (and maybe sparc-leon-linux if we want one).
--
Eric Botcazou
FDIV? 16 and FSQRT? 24 cycles,
> non-pipelined on separate unit
> GR FPU Lite: 8/8, with FDIVS/FDIVD/FSQRTS/FSQRTD 31/57/46/57 cycles,
> non-pipelined on same unit
Let's not make this too complex for a first try, the settings used at AdaCore
use [-mcpu=v8 / -msoft-float ]
>at compile-time to select the hardware
> setting.
--with-cpu shouldn't change multilibs. Multilibs are a property of a target,
e.g. sparc-leon-elf or sparc-rtems, not that of a cpu setting.
--
Eric Botcazou
Include leon.md scheduling description.
* config/sparc/leon.md: New file.
* config/sparc/t-elf: Do not assemble Solaris startup files.
* config/sparc/t-leon: New file.
* config/sparc/t-leon3: Likewise.
--
Eric Botcazou
Index: doc/invoke.texi
==
s for the ChangeLog.
--
Eric Botcazou
s during stage3? This will give no benefits
to the users and may introduce new bugs; we already have enough of them and
we should concentrate on fixing them. IMO this is stage1 material only.
--
Eric Botcazou
> Any idea why these results would be so poor ?
Because of your bootstrap settings, more precisely --enable-checking=all.
--
Eric Botcazou
> I was wondering about that lately. Should testsuite failures with
> --enable-checking=all be reported? IIRC, the 4.5 branch won't even
> bootstrap with that setting.
I'd think so, but only for the trunk probably.
--
Eric Botcazou
> So he updated his tree, saw changes in the middle-end and committed
> his without testing.
So Kenny would have had to lauch a new bootstrap, wait for a couple of hours
only to discover that something again changed in-between, and so on?
--
Eric Botcazou
> I don't know whether it would take him hours, since the tree does not
> even bootstrap, but most certainly Zdenek's statement was accurate and
> our commit procedure wasn't observed.
I'm not sure the commit procedure requires you to retest in that case.
--
Eric Botcazou
> Hm, then sol2.h and sol26.h build the minority where we have pthread_s_.
> Don't know the history yet.
That doesn't seem to be a Sun-ism.
--
Eric Botcazou
pass either
-static-libgcc or -shared-libgcc to link the EH machinery. That seems weird.
Geoff, any chance to bring Darwin back on par with the other platforms?
--
Eric Botcazou
nally can theoritically
cause problems on Linux when C code using POSIX threads is linked. So we
would need to do that only on Darwin.
--
Eric Botcazou
guess we have no other solution than defaulting to -static-libgcc for
Ada on Darwin.
Thanks for your feedback.
--
Eric Botcazou
dencies for "future binary
compatibility", but I see your point. I'll let Arnaud decide.
Thanks again.
--
Eric Botcazou
you can reproduce it after cleaning up the Solaris 9 machine.
--
Eric Botcazou
pected failure on all platforms (PR ada/20548).
Thanks for testing the Ada compiler!
--
Eric Botcazou
> Compile GMP 4.1.4:
>
> $ ../configure ABI=32 --prefix=/gcc-test --enable-mpfr
Would you mind trying as documented in
http://gcc.gnu.org/install/specific.html#sparc-sun-solaris2
i.e. with --build=sparc-sun-solaris2.9 instead of ABI=32?
Thanks in advance.
--
Eric Botcazou
ibraries, the processor itself, or
> any combination). The bug ID is 25998 for those interested.
I'm not sure we'll be able to sort it out but, in any case, thanks for
narrowing down the problem and for the extensive testing.
--
Eric Botcazou
> Could someone with powerpc-darwin access check wether Ada is now working
> or not, if not populate bugzilla?
The compiler bootstraps fine but all ACATS tests fail to link. You need to
manually pass -static-libgcc at link time.
--
Eric Botcazou
> Note that I opened PR ada/26296: "cxg2007 cxg2012 verify_flow_info
> failed" I'm seeing on x86_64-linux, may be something was recently broken
> on exception handling code and this could show up differently on hppa.
No, that's a different problem.
--
Eric Botcazou
.
>
> Is the patch only for ada? Will there be a C/C++-Version available
> someday?
The patch was developed for 3.x but, as per the GCC rules, needs to be
ported to 4.x and merged in mainline. We are in the process of doing that.
--
Eric Botcazou
his point, there nothing more we
can check ourselves here. Look at the base type if we have one and it
has the same precision. */
if (TREE_CODE (type) == INTEGER_TYPE
&& TREE_TYPE (type) != 0
&& TYPE_PRECISION (type) == TYPE_PRECISION (TREE_TYPE (type)
> http://gcc.gnu.org/ml/gcc-testresults/2006-02/msg01095.html
FAIL: cc1010b
Artifact or real failure?
--
Eric Botcazou
> One of the usual artifacts ...
Thanks. I've personally got them on Linux only, never ever on Solaris.
--
Eric Botcazou
we really care about it for Doug's work.
Thanks for your feedback.
--
Eric Botcazou
s about the bounds of types, but it does
virtually use the whole spectrum of TYPE_PRECISION, TYPE_MAX_VALUE and
TYPE_MIN_VALUE settings, unlike the C-family of front-ends.
This problem was already raised when Diego contributed the VRP pass and Diego
ajusted it to cope with Ada. AFAIK Ada and VRP work fine on the 4.1 branch.
--
Eric Botcazou
PE_UNSIGNED? How do you represent subtypes with non-canonical bounds?
--
Eric Botcazou
uble? */
I think we would need to clarify that, because I'm not sure the Ada front-end
directly generates:
D.1480_32 = nam_30 - 30361;
if (D.1480_32 <= 1) goto ; else goto ;
:;
D.1480_94 = ASSERT_EXPR ;
--
Eric Botcazou
GNU tools), probably an assembler problem/limitation.
--
Eric Botcazou
write the patch, provided that
an adept at the new bootstrap (black) magic gives me a clue as to where I
should start. :-)
--
Eric Botcazou
rmance anyway, so
> having a compliant implementation by default won't harm.
Changing compilation options to paper over potential bugs is IMHO not the best
approach, especially if the new options are little used. And the current
implementation of -gnato makes it a "heavy" option, not a "light" one.
--
Eric Botcazou
> it's not a bug, -gnato is clearly documented as required in this
> case, what makes you think otherwise?
Laurent's message.
Sorry about that, -gnato indeed has always been specified for this test.
--
Eric Botcazou
wn to the
front-end. */
#define TYPE_EXTRA_SUBTYPE_P(NODE) TYPE_LANG_FLAG_2 (NODE)
c460008__unsigned_edge_8 has the TYPE_EXTRA_SUBTYPE_P flag set.
--
Eric Botcazou
_LANG_FLAG_2 (NODE)
>
> c460008__unsigned_edge_8 has the TYPE_EXTRA_SUBTYPE_P flag set.
Which is of course a bit confusing because get_ada_base_type precisely returns
the first subtype "not known to the front-end"...
--
Eric Botcazou
hecking in that change.
Right, we need to fix it ASAP because a testsuite hang on x86 is pretty much a
show stopper. Richard, could you please install the change?
--
Eric Botcazou
> I missed the fact that the test was already in overflow.lst :)
No worries, so did I. :-)
--
Eric Botcazou
stage1_cflags="${stage1_cflags} -fkeep-inline-functions"
>fi
>
> since most versions of GCC (at least 2.7.2 which is the oldest I ever
> used, around 1998 on RH5.2) have the flag.
I'm not an autoconf guru but this test is trivial to write so I'd rather play
by the most strict rules.
--
Eric Botcazou
html
http://gcc.gnu.org/ml/gcc-testresults/2006-03/msg00345.html
http://gcc.gnu.org/ml/gcc-testresults/2006-03/msg00344.html
http://gcc.gnu.org/ml/gcc-testresults/2006-03/msg00343.html
http://gcc.gnu.org/ml/gcc-testresults/2006-03/msg00342.html
--
Eric Botcazou
1, build_int_cst (etype, 0), value);
}
I gather that we should restrict the transformation to INTEGER_TYPEs.
--
Eric Botcazou
their base type.
I'm testing a pure fold-const.c patch based on Richard's suggestion.
--
Eric Botcazou
> One more note, we see the same kind of conditional and test
> simplification with for cxa4028 in Ada.Strings.Superbounded.Super_Trim.
> So I'm pretty confident that if we fix the bogus trees generated for
> a-stwifi.adb that all three of these regressions will be fixed.
Con
so "GCC for SPARC Systems" is a bit
misleading, given that FSF GCC for SPARC does run on the aforementioned
operating systems in addition to Solaris. Something like "Sun GCC for
SPARC/Solaris Systems" although I'm not sure if using "GCC" is not already
misleading.
--
Eric Botcazou
> GCC 4.0.3 has been released.
You need to add a link on the page
http://gcc.gnu.org/gcc-4.0
Similarly for the 4.1.0 release on the page
http://gcc.gnu.org/gcc-4.1
And don't you need to display the release date instead of "current changes"
for 4.1.0 on the main page?
--
Eric Botcazou
,
> so at the same time it would be good to change it to v8plus as well.
>
> All that changes can be done by easily tweaking gcc/config.gcc
> by adding with_cpu=v9
I'm all for it, at least for Solaris 7+.
--
Eric Botcazou
> It seems everybody agreed that solaris 10+ can be changed to -mcpu=v9
> default. Great!
> What are the thoughts about Solaris 7,8,9 ?
Go ahead, post the patch for Solaris 7+ on gcc-patches@, I'll install it.
--
Eric Botcazou
> Hi Jeff, this seems to work nicely - thanks again.
Well, this has introduced 3 regressions in the ACATS testsuite on x86/x86-64.
--
Eric Botcazou
> Which ones?
PR tree-optimization/26797.
--
Eric Botcazou
601 - 700 of 1095 matches
Mail list logo
