Handle return values from fchown

The function fchown is annotated with warn_unused_result.  As
formerly used in the code, it would emit a compiler warning
```warning: ignoring return value of ‘fchown’, declared with
attribute warn_unused_result [-Wunused-result]```

This commit notes the return value and emits appropriate error/logging
messages if the call fails, creating more traceable results and
satisfying the compiler.
This commit is contained in:
Jeff Kowalski 2016-03-12 11:21:52 -08:00
parent 540bdfcb02
commit 46b9f263ac
2 changed files with 9 additions and 5 deletions

View file

@ -859,22 +859,23 @@ bool env_universal_t::sync(callback_data_list_t *callbacks)
success = this->write_to_fd(private_fd, private_file_path);
if (! success) UNIVERSAL_LOG("write_to_fd() failed");
}
if (success)
{
/* Ensure we maintain ownership and permissions (#2176) */
struct stat sbuf;
if (wstat(vars_path, &sbuf) >= 0)
{
fchown(private_fd, sbuf.st_uid, sbuf.st_gid);
if (0 > fchown(private_fd, sbuf.st_uid, sbuf.st_gid))
UNIVERSAL_LOG("fchown() failed");
fchmod(private_fd, sbuf.st_mode);
}
/* Linux by default stores the mtime with low precision, low enough that updates that occur in quick succession may
result in the same mtime (even the nanoseconds field). So manually set the mtime of the new file to a high-precision
clock. Note that this is only necessary because Linux aggressively reuses inodes, causing the ABA problem; on other
platforms we tend to notice the file has changed due to a different inode (or file size!)
It's probably worth finding a simpler solution to this. The tests ran into this, but it's unlikely to affect users.
*/
#if HAVE_CLOCK_GETTIME && HAVE_FUTIMENS

View file

@ -1453,7 +1453,10 @@ bool history_t::save_internal_via_rewrite()
if (wstat(new_name, &sbuf) >= 0)
{
/* Success */
fchown(out_fd, sbuf.st_uid, sbuf.st_gid);
if (0 > fchown(out_fd, sbuf.st_uid, sbuf.st_gid))
{
debug(2, L"Error when changing ownership of history file");
}
fchmod(out_fd, sbuf.st_mode);
}