[PATCH] Run the resolver in parallel with the reader.

Rui Ueyama ruiu at google.com
Thu Jan 15 01:12:48 PST 2015

Hi Bigcheese, kledzik, atanasyan, shankarke, Jean-Daniel,

This patch makes File::parse() multi-thread safe. If one thread is running
File::parse(), other threads will block if they try to call the same method.
File::parse() is idempotent, so you can safely call  multiple times.

With this change, we don't have to wait for all worker threads to finish
in Driver::link(). Previously, Driver::link() calls TaskGroup::sync() to
wait for all threads running File::parse(). This was not ideal because
we couldn't start the resolver until we parse all files.

This patch increase parallelism by making Driver::link() to not wait for
worker threads. The resolver calls parse() to make sure that the file
being read has been parsed, and then uses the file. In this approach,
the resolver can run with the parser threads in parallel.

(This patch also opens up a possibility to read files from archive files
speculatively. Currently, reading member files from an archive file is
pretty slow because it's serialized, since the resolver is a single-
threaded routine. We could speed this thing up by instantiating member
files that are likely to be needed. As long as a file instantiation and
its parse() method are side-effect free, that operation does not affect
correctness. I don't do that in this patch, but will try in future.)



-------------- next part --------------
A non-text attachment was scrubbed...
Name: D6994.18217.patch
Type: text/x-patch
Size: 8232 bytes
Desc: not available
URL: <http://lists.llvm.org/pipermail/llvm-commits/attachments/20150115/08d4cd3d/attachment.bin>

More information about the llvm-commits mailing list