Uh oh!
There was an error while loading. Please reload this page.
GH-141794: Limit size of generated machine code. - #142228
Conversation
* Factor out bodies of the largest uops, to reduce jit code size. * Factor out common assert, also reducing jit code size. * Limit size of jitted code for a single executor to 1MB.
diegorusso
left a comment
There was a problem hiding this comment.
Left a couple of minor comments.
Uh oh!
There was an error while loading. Please reload this page.
Uh oh!
There was an error while loading. Please reload this page.
Co-authored-by: Diego Russo <diego.russo@arm.com>
Uh oh!
There was an error while loading. Please reload this page.
bedevere-bot
commented
Dec 3, 2025
|
bedevere-bot
commented
Dec 3, 2025
|
Fidget-Spinner
left a comment
There was a problem hiding this comment.
This has a bunch of refleaks which I've yet to find, but here are some easier ones to spot.
| res = NULL; | ||
| goto cleanup; | ||
| } | ||
| res = _PyUnicode_JoinArray(&_Py_STR(empty), args_o, total_args); |
There was a problem hiding this comment.
No stackref buffer cleanup here.
| args_o, 2, | ||
| args_o+1, 2, | ||
| half_args | ||
| ); |
There was a problem hiding this comment.
No stackref buffer cleanup here.
markshannon
commented
Dec 4, 2025
Refleaks fixed in #142257 |
mhsmith
commented
Dec 4, 2025
It looks like these leaks may also have been the cause of this failure in the Android CI, which appeared when this PR was merged and disappeared when #142257 was merged. However, this was confused by a different Android failure which was happening around the same time. |
* Factor out bodies of the largest uops, to reduce jit code size. * Factor out common assert, also reducing jit code size. * Limit size of jitted code for a single executor to 1MB.
Uh oh!
There was an error while loading. Please reload this page.