Uh oh!
There was an error while loading. Please reload this page.
Remove check-fast. Closes #4193, #8844, #6330, #7416 - #12304
Conversation
huonw
commented
Feb 16, 2014
Out of interest, how long does the full test suite take to run on windows? (Also, needs a rebase.) |
alexcrichton
commented
Feb 16, 2014
I, too, am curious about how this would affect cycle time. We definitely don't want cycle time to go up, so I think we should see how long the suite is on windows before merging. |
brson
commented
Feb 16, 2014
I can run a test. |
brson
commented
Feb 16, 2014
On win1 before removal check-fast took 16m; after, check-lite takes 24m. |
alexcrichton
commented
Feb 16, 2014
That doesn't sound nearly as bad as I thought it was, but it also may push the windows bots to being the slowest bots. I think that the linux bots would be faster by 10-20min at that point. That doesn't sound too bad, and we could always slim down slower portions of the windows test suite if necessary. |
adrientetar
commented
Feb 16, 2014
It sounds like we could largely speed-up Windows according to #8844 (comment). |
brson
commented
Feb 16, 2014
If check-lite is not fast enough we can change it to test fewer things. |
alexcrichton
commented
Mar 10, 2014
For the compile-fail test, is it possible to do If so, r=me |
This fixes struct passing abi on x86 ffi: Structs are now passed indirectly with byval attribute (as clang does).
This should make BSD use the proper GNU make.
alexcrichton
commented
Apr 3, 2014
Closing in favor of #13288 |
…ents, r=jonas-schievink minor: Include self type in generated getter/setter docs
No description provided.