[distcc] small protocol extensions for -ftest-coverage, -fbranch-probabilities

[Martin Pool]

On 24 Jun 2004, Dan Kegel <dank at kegel.com> wrote:

> I guess you'd need to add an OUFN to the request to give the
> name of each output file, then.  And the DOTO would be replaced
> with just a generic OFIL token, I suppose.

Yes, and similarly coming back.  I just didn't write that bit yet.

Does that sound OK?

> BTW we'll have to change the server to use the proper .i name,
> presumably in a randomly-named directory, otherwise the .da files
> are generated with the wrong name.

Yes, one thing I wanted to do was to allow the client to control the
input and output filenames, which implies a clean directory for each
job.

I think this will make it easier to support tasks other than gcc:
either different compilers or something other than C compilation.

> (And I have to patch gcc anyway to set the directory name used for
> the .da files.  Maybe I could get away with combining the two, but I
> don't think so.)

Could you explain more about this?

-- 
Martin 

-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: application/pgp-signature
Size: 189 bytes
Desc: Digital signature
Url : http://lists.samba.org/archive/distcc/attachments/20040625/260bd4b6/attachment.bin