[PATCH] D21393: [pdbdump] Verify LF_{CLASS, ENUM, INTERFACE, STRUCTURE, UNION} records with different hash function.

Zachary Turner via llvm-commits llvm-commits at lists.llvm.org
Wed Jun 15 12:40:38 PDT 2016

Ahh, I think I get it.  I'm a little concerned about the performance impact
of this though.  The whole reason the hash is stored in the file is so that
it doesn't have to recompute the hash up front.  If we're re-computing it
every time we load the type streams, it seems like it will be a big
performance problem.  I think we should only compute the hash when we
modify a record and/or write the file.

On Wed, Jun 15, 2016 at 12:02 PM Rui Ueyama <ruiu at google.com> wrote:

> ruiu added a comment.
> I think that's exactly what we are doing -- in this code, we read type
> records, compute hash values using our hash function, and compare them with
> hash values on the PDB file. If they are different, pdbdump reports that
> the file is corrupted.
> http://reviews.llvm.org/D21393
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.llvm.org/pipermail/llvm-commits/attachments/20160615/a31d6725/attachment.html>

More information about the llvm-commits mailing list