Displaying 5 results from an estimated 5 matches for "patch_without".
2013 Oct 17
1
[LLVMdev] [Debug Info PATCH] for support of ref_addr and removal of DIE duplication
...gt;>
>> Patches are reattached for convenience: remove DIE duplication with a
>> worklist (name it patch_with), remove DIE duplication without a worklist
>> but an assertion when we emit a ref4, we make sure the DIE and the
>> referenced DIE belong to the same CU (name it patch_without).
>>
>
> OK, we're going round in circles here and I'm not sure there are many
> other ways I can communicate things.
>
>
>> Without my patch, the assumption may be true, but it does not matter
>> since we should always use ref4.
>> I have provided som...
2013 Oct 17
2
[LLVMdev] [Debug Info PATCH] for support of ref_addr and removal of DIE duplication
...;t given it a great deal of thought.
>
Patches are reattached for convenience: remove DIE duplication with a
worklist (name it patch_with), remove DIE duplication without a worklist
but an assertion when we emit a ref4, we make sure the DIE and the
referenced DIE belong to the same CU (name it patch_without).
Without my patch, the assumption may be true, but it does not matter since
we should always use ref4.
I have provided some cases that the assertion fails with patch_without.
I didn't get a chance to implement another assertion we mentioned earlier
(verify that inside addDIEEntry if a DIE do...
2013 Oct 17
0
[LLVMdev] [Debug Info PATCH] for support of ref_addr and removal of DIE duplication
...ght.
>>
>
>
> Patches are reattached for convenience: remove DIE duplication with a
> worklist (name it patch_with), remove DIE duplication without a worklist
> but an assertion when we emit a ref4, we make sure the DIE and the
> referenced DIE belong to the same CU (name it patch_without).
>
OK, we're going round in circles here and I'm not sure there are many other
ways I can communicate things.
> Without my patch, the assumption may be true, but it does not matter since
> we should always use ref4.
> I have provided some cases that the assertion fails with...
2013 Oct 16
0
[LLVMdev] [Debug Info PATCH] for support of ref_addr and removal of DIE duplication
On Wed, Oct 16, 2013 at 1:21 PM, Manman Ren <manman.ren at gmail.com> wrote:
>
> There are a few places where we break the assumption:
> 1> formal_parameter constructed in DwarfDebug when adding attribute type
> we call SPCU->addType(Arg, ATy), where Arg does not belong to SPCU.
> 2> inlined_subroutine constructed in DwarfDebug when adding attribute
>
2013 Oct 16
2
[LLVMdev] [Debug Info PATCH] for support of ref_addr and removal of DIE duplication
There are a few places where we break the assumption:
1> formal_parameter constructed in DwarfDebug when adding attribute type
we call SPCU->addType(Arg, ATy), where Arg does not belong to SPCU.
2> inlined_subroutine constructed in DwarfDebug when adding attribute
abstract_origin
The inlined_subroutine does not belong to the CU we call addDIEEntry
on.
We create the children