[PATCH] DebugInfo: Remove 'inlinedAt:' field from MDLocalVariable/DIVariable

[David Blaikie]

On Apr 1, 2015 11:36 PM, "Duncan P. N. Exon Smith" <dexonsmith at apple.com>
wrote:
>
> When I was gathering LTO heap profiles a few weeks ago, the single
> largest `Metadata` contributer to memory usage was `MDLocalVariable`,
> with a huge chunk coming from `MDLocalVariable::withInline()`.
>
> This patch stops storing where an `MDLocalVariable` has been inlined.
> This inlined-at info was used two ways:
>
>   - To tell the backend where a variable was inlined.
>   - To create a unique id for each inlined variable.
>
> However, every instruction has a `!dbg` attachment with the same
> information in its inlined-at field.  AFAICT, David Blaikie's work to
> make that accurate has paid off, since this naive patch (which just
> removes the field entirely) passes all the tests.
>
> The unique id for inlined variables is now a typedef called
> `InlinedVariable` of `std::pair<MDLocalVariable*, MDLocation*>`.
>
> Besides the patch, I've attached the upgrade script I used to update
> testcases.
>
> Any reason not to commit?  (Am I missing something here?)

Last time I tried this (not sure whether I got as far as
commuting/referring it) either I hit those issues I have since fixed (maybe
that's part of what me going on that kick) or some other issues...

In any case, would it be possible to run some tests that assert that these
two locations are the same (possibly this assertion could take the place of
the construction of these excessive/redundant/bloated inlined DIVariables -
so we could gain the memory savings while still leaving the checking in
just in case there are new/weird/interesting cases) for now?

>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.llvm.org/pipermail/llvm-commits/attachments/20150402/e8f93ca2/attachment.html>