mprintf: do not ignore length modifiers of %o, %x, %X - #15348
Conversation
There are uses of `%lx` and `%zx` in the codebase, but `parsefmt` interpreted them as `%x`.
|
Analysis of PR #15348 at 22423c41: Test 284 failed, which has NOT been flaky recently, so there could be a real issue in this PR. Test 285 failed, which has NOT been flaky recently, so there could be a real issue in this PR. Test 980 failed, which has NOT been flaky recently, so there could be a real issue in this PR. Test 1049 failed, which has NOT been flaky recently, so there could be a real issue in this PR. Test 1093 failed, which has NOT been flaky recently, so there could be a real issue in this PR. Test 1094 failed, which has NOT been flaky recently, so there could be a real issue in this PR. Test http/test_09_push.py::TestPush::test_09_02_h2_push[0] failed, which has NOT been flaky recently, so there could be a real issue in this PR. There are more failures, but that's enough from Gha. Generated by Testclutch |
|
Lovely! Do you think test 557 can be expanded to test %lx and %zx as well ? |
Also |
Please do, and I figure getting just something added there will be a step up from having nothing!
Sounds appropriate. %o should be considered unsigned! |
`%x` and `%X` were already treated as unsigned, but `%o` was not, even though it was used with unsigned numbers.
`%x` and `%X` were already treated as unsigned, but `%o` was not, even though it was used with unsigned numbers. Closes #15348
|
Thanks! |
There are uses of `%lx` and `%zx` in the codebase, but `parsefmt` interpreted them as `%x`. Closes curl#15348
`%x` and `%X` were already treated as unsigned, but `%o` was not, even though it was used with unsigned numbers. Closes curl#15348
3 participants
There are uses of
%lxand%zxin the codebase, butparsefmtinterpreted them as%x.