This is an archive of the discontinued LLVM Phabricator instance.

[SimplifyLibCalls] Improve double->float shrinking to consider constants
ClosedPublic

Authored by MatzeB on Dec 2 2014, 7:38 PM.

Diff Detail

Event Timeline

MatzeB updated this revision to Diff 16848.Dec 2 2014, 7:38 PM
MatzeB retitled this revision from to [SimplifyLibCalls] Improve double->float shrinking to consider constants.
MatzeB updated this object.
MatzeB edited the test plan for this revision. (Show Details)
MatzeB added a subscriber: Unknown Object (MLST).
qcolombet accepted this revision.Dec 3 2014, 10:11 AM
qcolombet added a reviewer: qcolombet.
qcolombet added a subscriber: qcolombet.

Hi Matthias,

LGTM.

Thanks,
-Quentin

This revision is now accepted and ready to land.Dec 3 2014, 10:11 AM
pete added a subscriber: pete.Dec 3 2014, 11:05 AM

This should't block committing, but could you please also add a test for a constant which does lose info when converted to float, and so fails this optimization.

Thanks,
Pete

MatzeB updated this revision to Diff 16879.Dec 3 2014, 11:32 AM
MatzeB edited edge metadata.

Added negative test where transformation should not be applied

Good point! I added another test.

pete accepted this revision.Dec 3 2014, 11:38 AM
pete added a reviewer: pete.

Thanks. LGTM.

Diffusion closed this revision.Dec 3 2014, 1:47 PM
Diffusion updated this revision to Diff 16889.

Closed by commit rL223270 (authored by matze).