[llvm] r219010 - Revert "Revert "DI: Fold constant arguments into a single MDString""

Adrian Prantl aprantl at apple.com
Fri Oct 3 17:37:02 PDT 2014


> On Oct 3, 2014, at 5:34 PM, Duncan P. N. Exon Smith <dexonsmith at apple.com> wrote:
> 
> Oops!  Updating that document was faster than I expected.  This must
> be why.
> 
> Looking more closely, the examples don't seem better than looking
> directly at `include/llvm/IR/DebugInfo.h` and `lib/IR/DIBuilder.cpp`.
> 
> In the attached patch, I update a few of them, but prune the rest, and
> point the reader at the appropriate place in the source.
> 
> Thoughts?

Looks like you found a way to reduce the maintenance effort while increasing the accuracy of the documentation. I won’t complain :-)

> 
> <0001-DI-Update-and-prune-metadata-examples.patch>
>> On Oct 3, 2014, at 3:51 PM, Adrian Prantl <aprantl at apple.com> wrote:
>> 
>> I just noticed that the update to docs/SourceLevelDebugging.rst misses the second part of the document were all the examples for C/C++ are listed.
>> 
>> -- adrian
>> 
>> 
>>> On Oct 3, 2014, at 1:01 PM, Duncan P. N. Exon Smith <dexonsmith at apple.com> wrote:
>>> 
>>> Author: dexonsmith
>>> Date: Fri Oct  3 15:01:09 2014
>>> New Revision: 219010
>>> 
>>> URL: http://llvm.org/viewvc/llvm-project?rev=219010&view=rev
>>> Log:
>>> Revert "Revert "DI: Fold constant arguments into a single MDString""
>>> 
>>> This reverts commit r218918, effectively reapplying r218914 after fixing
>>> an Ocaml bindings test and an Asan crash.  The root cause of the latter
>>> was a tightened-up check in `DILexicalBlock::Verify()`, so I'll file a
>>> PR to investigate who requires the loose check (and why).
>>> 
>>> Original commit message follows.
>>> 
>>> --
>>> 
>>> This patch addresses the first stage of PR17891 by folding constant
>>> arguments together into a single MDString.  Integers are stringified and
>>> a `\0` character is used as a separator.
>>> 
>>> Part of PR17891.
>>> 
>>> Note: I've attached my testcases upgrade scripts to the PR.  If I've
>>> just broken your out-of-tree testcases, they might help.
> 





More information about the llvm-commits mailing list