Clang warns:
../fs/f2fs/data.c:995:17: error: implicit truncation from 'int' to a one-bit wide bit-field changes value from 1 to -1 [-Werror,-Wsingle-bit-bitfield-constant-conversion]
fio->submitted = 1;
^ ~
../fs/f2fs/data.c:1011:15: error: implicit truncation from 'int' to a one-bit wide bit-field changes value from 1 to -1 [-Werror,-Wsingle-bit-bitfield-constant-conversion]
fio->retry = 1;
^ ~
../fs/f2fs/segment.c:3320:16: error: implicit truncation from 'int' to a one-bit wide bit-field changes value from 1 to -1 [-Werror,-Wsingle-bit-bitfield-constant-conversion]
fio->in_list = 1;
^ ~
There is not a bug here because the value of these fields is never
explicitly compared against (just whether it is zero or non-zero) but
it is easy to silence the warning by using an unsigned type to allow
an assignment of 0 or 1 without implicit conversion.
Fixes: 998863dadd2c ("f2fs: reduce stack memory cost by using bitfield in struct f2fs_io_info")
Link: https://github.com/ClangBuiltLinux/linux/issues/1796
Reported-by: kernel test robot <[email protected]>
Signed-off-by: Nathan Chancellor <[email protected]>
---
fs/f2fs/f2fs.h | 12 ++++++------
1 file changed, 6 insertions(+), 6 deletions(-)
diff --git a/fs/f2fs/f2fs.h b/fs/f2fs/f2fs.h
index 08dc64c5050e..89f6fdfeed19 100644
--- a/fs/f2fs/f2fs.h
+++ b/fs/f2fs/f2fs.h
@@ -1213,12 +1213,12 @@ struct f2fs_io_info {
int compr_blocks; /* # of compressed block addresses */
int need_lock:8; /* indicate we need to lock cp_rwsem */
int version:8; /* version of the node */
- int submitted:1; /* indicate IO submission */
- int in_list:1; /* indicate fio is in io_list */
- int is_por:1; /* indicate IO is from recovery or not */
- int retry:1; /* need to reallocate block address */
- int encrypted:1; /* indicate file is encrypted */
- int post_read:1; /* require post read */
+ unsigned int submitted:1; /* indicate IO submission */
+ unsigned int in_list:1; /* indicate fio is in io_list */
+ unsigned int is_por:1; /* indicate IO is from recovery or not */
+ unsigned int retry:1; /* need to reallocate block address */
+ unsigned int encrypted:1; /* indicate file is encrypted */
+ unsigned int post_read:1; /* require post read */
enum iostat_type io_type; /* io type */
struct writeback_control *io_wbc; /* writeback control */
struct bio **bio; /* bio for ipu */
---
base-commit: de6b3a5e09b29c014bd04044b023896107cfa2ee
change-id: 20230201-f2fs-fix-single-length-bitfields-df8cc78e880a
Best regards,
--
Nathan Chancellor <[email protected]>
On 2023/2/2 0:40, Nathan Chancellor wrote:
> Clang warns:
>
> ../fs/f2fs/data.c:995:17: error: implicit truncation from 'int' to a one-bit wide bit-field changes value from 1 to -1 [-Werror,-Wsingle-bit-bitfield-constant-conversion]
> fio->submitted = 1;
> ^ ~
> ../fs/f2fs/data.c:1011:15: error: implicit truncation from 'int' to a one-bit wide bit-field changes value from 1 to -1 [-Werror,-Wsingle-bit-bitfield-constant-conversion]
> fio->retry = 1;
> ^ ~
>
> ../fs/f2fs/segment.c:3320:16: error: implicit truncation from 'int' to a one-bit wide bit-field changes value from 1 to -1 [-Werror,-Wsingle-bit-bitfield-constant-conversion]
> fio->in_list = 1;
> ^ ~
>
> There is not a bug here because the value of these fields is never
> explicitly compared against (just whether it is zero or non-zero) but
> it is easy to silence the warning by using an unsigned type to allow
> an assignment of 0 or 1 without implicit conversion.
Nathan, thanks a lot for catching this, do you mind letting I merge this fix
into original patch? as the original patch has not been upstreamed yet.
Thanks,
On Thu, Feb 02, 2023 at 02:13:18PM +0800, Chao Yu wrote:
> On 2023/2/2 0:40, Nathan Chancellor wrote:
> > Clang warns:
> >
> > ../fs/f2fs/data.c:995:17: error: implicit truncation from 'int' to a one-bit wide bit-field changes value from 1 to -1 [-Werror,-Wsingle-bit-bitfield-constant-conversion]
> > fio->submitted = 1;
> > ^ ~
> > ../fs/f2fs/data.c:1011:15: error: implicit truncation from 'int' to a one-bit wide bit-field changes value from 1 to -1 [-Werror,-Wsingle-bit-bitfield-constant-conversion]
> > fio->retry = 1;
> > ^ ~
> >
> > ../fs/f2fs/segment.c:3320:16: error: implicit truncation from 'int' to a one-bit wide bit-field changes value from 1 to -1 [-Werror,-Wsingle-bit-bitfield-constant-conversion]
> > fio->in_list = 1;
> > ^ ~
> >
> > There is not a bug here because the value of these fields is never
> > explicitly compared against (just whether it is zero or non-zero) but
> > it is easy to silence the warning by using an unsigned type to allow
> > an assignment of 0 or 1 without implicit conversion.
>
> Nathan, thanks a lot for catching this, do you mind letting I merge this fix
> into original patch? as the original patch has not been upstreamed yet.
No worries, do whatever you need to! I just care that the problem gets
resolved one way or another :)
Cheers,
Nathan
On 2023/2/2 14:15, Nathan Chancellor wrote:
> On Thu, Feb 02, 2023 at 02:13:18PM +0800, Chao Yu wrote:
>> On 2023/2/2 0:40, Nathan Chancellor wrote:
>>> Clang warns:
>>>
>>> ../fs/f2fs/data.c:995:17: error: implicit truncation from 'int' to a one-bit wide bit-field changes value from 1 to -1 [-Werror,-Wsingle-bit-bitfield-constant-conversion]
>>> fio->submitted = 1;
>>> ^ ~
>>> ../fs/f2fs/data.c:1011:15: error: implicit truncation from 'int' to a one-bit wide bit-field changes value from 1 to -1 [-Werror,-Wsingle-bit-bitfield-constant-conversion]
>>> fio->retry = 1;
>>> ^ ~
>>>
>>> ../fs/f2fs/segment.c:3320:16: error: implicit truncation from 'int' to a one-bit wide bit-field changes value from 1 to -1 [-Werror,-Wsingle-bit-bitfield-constant-conversion]
>>> fio->in_list = 1;
>>> ^ ~
>>>
>>> There is not a bug here because the value of these fields is never
>>> explicitly compared against (just whether it is zero or non-zero) but
>>> it is easy to silence the warning by using an unsigned type to allow
>>> an assignment of 0 or 1 without implicit conversion.
>>
>> Nathan, thanks a lot for catching this, do you mind letting I merge this fix
>> into original patch? as the original patch has not been upstreamed yet.
>
> No worries, do whatever you need to! I just care that the problem gets
> resolved one way or another :)
Thank you! Updated a v3 patch. :)
Thanks,
>
> Cheers,
> Nathan