Re: [PATCH x86, java, PR63536] Fix java bootstrap for -mtune=intel/slm

Evgeny Stupachenko <[email protected]> Wed, 15 Oct 2014 21:12:13 +0400
Newsgroups gmane.comp.gcc.java.patches
Message-ID <CAOvf_xyotqg3w-nd8LJqswBvQWQnOgmC=jmq2cwFv3f6BMzTkw@mail.gmail.com>
Updated ChangeLog:

2014-10-15  Evgeny Stupachenko  <[email protected]>

        PR target/63536
        * lang.c (java_print_error_function): Add check on NULL function
        context.

The function is: _Jv_global_static_constructor()
It is compiler generated at jcf-parse.c
(java_emit_static_constructor), line 1705.

On Wed, Oct 15, 2014 at 8:59 PM, Jakub Jelinek <[email protected]> wrote:
> On Wed, Oct 15, 2014 at 05:57:22PM +0100, Andrew Haley wrote:
>> On 10/15/2014 05:54 PM, Evgeny Stupachenko wrote:
>> > The patch fixes java i686 bootstrap for -mtune=intel/slm.
>> >
>> > Recent changes triggered java to write a note on compilation for a
>> > function without context.
>> >
>> > make check in progress
>> >
>> > Is it ok?
>>
>> I guess so, but I don't understand how any function ends up with a null
>> DECL_CONTEXT.
>
> Yeah, I guess it would help if you would say what function it is exactly,
> how it got created etc.  Artificial function?
>
>> Nontheless, this must go to [email protected]
>
> Also, java/ prefix shouldn't be in the ChangeLog, gcc/java/ directory has
> its own ChangeLog.
>
>         Jakub