Repository navigation
Conversation
ac32567 to
802f51b
Compare
|
I think the fix is correct, but do not think it qualifies for 8.2 ? |
ndossche
left a comment
There was a problem hiding this comment.
Fix is right, see nits.
Indeed cannot qualify for a security branch.
| /* get the value */ | ||
| switch (sqlite3_value_type(argv[i])) { | ||
| case SQLITE_INTEGER: | ||
| #if ZEND_LONG_MAX > 2147483647 |
There was a problem hiding this comment.
Should use #ifdef ZEND_ENABLE_ZVAL_LONG64 instead (below too)
| @@ -0,0 +1,28 @@ | |||
| --TEST-- | |||
| GH-25021 UDF isn't using sqlite3_*_int64 API | |||
There was a problem hiding this comment.
| GH-25021 UDF isn't using sqlite3_*_int64 API | |
| GH-25021 (UDF isn't using sqlite3_*_int64 API) |
There was a problem hiding this comment.
Some names are formatted this way, some are not, why?
There was a problem hiding this comment.
because people are inconsistent
| @@ -0,0 +1,27 @@ | |||
| --TEST-- | |||
| GH-25021 UDF isn't using sqlite3_*_int64 API | |||
There was a problem hiding this comment.
| GH-25021 UDF isn't using sqlite3_*_int64 API | |
| GH-25021 (UDF isn't using sqlite3_*_int64 API) |
Even if it corrupts data and no user workaround is possible? |
Yes. Same with many other bugfixes |
fix #25021