This got changed to a double but all the callers still use a u64, which causes
us to segfault sometimes because of some weird C voodoo that I had to have
explained to me. Apparently because we're using a double the compiler will use
the floating point registers to hold our argument which ends up not being
aligned properly if you don't actually give it a double so it will cause
problems for other things, in our case it was screwing up str_bytes so it was
larger than the actual size of the str. This patch fixes the segfault. Thanks,
Signed-off-by: Josef Bacik <jbacik@fusionio.com>
Signed-off-by: Chris Mason <chris.mason@fusionio.com>
}
static char *size_strs[] = { "", "KiB", "MiB", "GiB", "TiB", "PiB", "EiB"};
-int pretty_size_snprintf(double size, char *str, size_t str_bytes)
+int pretty_size_snprintf(u64 size, char *str, size_t str_bytes)
{
int num_divs = 0;
float fraction;
int btrfs_device_already_in_root(struct btrfs_root *root, int fd,
int super_offset);
-int pretty_size_snprintf(double size, char *str, size_t str_bytes);
+int pretty_size_snprintf(u64 size, char *str, size_t str_bytes);
#define pretty_size(size) \
({ \
static __thread char _str[24]; \