> This patch causes an ICE when building libobjc, which is part of
> normal bootstrap.
>
> PR 58179
>
> /nasfarm/edelsohn/src/src/libobjc/accessors.m: In function 'objc_getProperty':
> /nasfarm/edelsohn/src/src/libobjc/accessors.m:127:11: internal
> compiler error: in obj_type_ref_class, at tree.
This patch from Chris Manghane fixes a bug when the value in a range
clause is a sink variable. It's fixedbugs/bug454.go in the master Go
testsuite, and will appear in the gccgo testsuite in due course.
Bootstrapped and ran Go testsuite on x86_64-unknown-linux-gnu.
Committed to mainline and 4.8 br
Hi,
this is the patch I comitted after testing.
My apologizes for the breakage.
Honza
Index: ChangeLog
===
--- ChangeLog (revision 201816)
+++ ChangeLog (working copy)
@@ -1,3 +1,8 @@
+2013-08-16 Jan Hubicka
+
+ PR middl
Hi,
I have got the ipa-devirt implementation into shape it seems actually useful
and working as expected. The main selling point of it for the moment is
compile time of C++ programs (especially with LTO) and verification facilities.
Currently we keep all virtual functions in the callgraph until a
>
> I added both of these and ran into issues due to profile maintenance.
> For example, there were non-zero blocks in the cold section because
> pro_and_epilogue split a simple return block that was previously reach
> by both hot and cold paths. The new return block that was then only
> reached v
On Sat, 17 Aug 2013, Jan Hubicka wrote:
> this is the patch I comitted after testing.
> My apologizes for the breakage.
>
> +2013-08-16 Jan Hubicka
> +
> + PR middle-end/58179
> + * tree.c (obj_type_ref_class): Do not ICE on non-method calls.
This fixes my bootstraps on {amd64,i386}-un
Hi,
this patch tries to make gimple_extract_devirt_binfo_from_cst little bit more
sane and make it match bit more cases. Currently we seem to be able to
devirtualize only if the type in question has no basetypes (not even some not
having virtual methods at all) and it is the initial type implement
On 08/17/2013 05:44 PM, Jan Hubicka wrote:
1) we want the type to not have base because we may have inlined the
constructor.
During construction the vtables are filled by base's vtable and thus we can
not simply devirtualize based on the final virtual table without proving
that
The standard says that lambdas go in the enclosing block, class or
namespace scope, but also that default arguments aren't parsed until the
class is complete. So we need to accept pushing a lambda into a
complete class.
Tested x86_64-pc-linux-gnu, applying to trunk and 4.8.
commit a3eed6be088
On Sat, Aug 17, 2013 at 7:56 PM, Jason Merrill wrote:
>> I do not know if one can do
>> something like having automatic variable of class A and use placement new
>> to change it to class B.
>
>
> This is something of a grey area in the standard, with a few defect reports
> yet to be resolved. I
In this testcase, the A conversion template can't possibly produce a
pointer type, so its presence in the overload set shouldn't cause an
error. The standard is unclear about the B case, but it seems
appropriate to me to still complain there.
Tested x86_64-pc-linux-gnu, applying to trunk, 4.8
Jason Merrill wrote:
>On 08/17/2013 05:44 PM, Jan Hubicka wrote:
>> 1) we want the type to not have base because we may have inlined
>the constructor.
>> During construction the vtables are filled by base's vtable and
>thus we can
>> not simply devirtualize based on the final virtual t
12 matches
Mail list logo
