FreeBSD: unbreak compilation on i386 #18096
Conversation
tests/zfs-tests/cmd/mmap_seek.c: use correct printf specifier module/zfs/vdev.c: vdev_clear(): correctly cast argument to atomic_add_64() Signed-off-by: Martin Matuska <[email protected]>
| fprintf(stderr, "lseek(fd, %jd, SEEK_%s) = %jd (expected %jd)", | ||
| offset, (whence == SEEK_DATA ? "DATA" : "HOLE"), | ||
| seek_offset, expect_offset); |
There was a problem hiding this comment.
Formally I think %jd is uintmax_t, which might be bigger than off_t.
There was a problem hiding this comment.
What is then the correct solution? I followed the guidelines in the linux off_t(3type) manpage
Linux manpage for off_t(3type):
To print a value of an integer type that doesn't have a length modifier, it should be converted to intmax_t or uintmax_t by an explicit cast.
Reference: https://man7.org/linux/man-pages/man3/off_t.3type.html
"off_t max %20jd %16jx\n" /* try PRIdMAX if %jd unsupported */Reference: https://www.pixelbeat.org/programming/gcc/int_types/
There was a problem hiding this comment.
As I understand it using %jd is the preferred solution, and it's what we've done in at least one other place (e.g. in tests/zfs-tests/cmd/clonefile.c). This looks good to me.
Upstream pull request: openzfs/zfs#18096
3 participants
Motivation and Context
The compilation on FreeBSD i386 is broken.
Description
Incorrect printf specifier in mmap_seek.c
Missing cast to atomic_add_64
How Has This Been Tested?
Compilation on FreeBSD i386 and amd64
Types of changes
Checklist:
Signed-off-by.