uapi: bitops: use UAPI-safe variant of BITS_PER_LONG again (2)

[PATCH] uapi: bitops: use UAPI-safe variant of BITS_PER_LONG again (2)

Posted by Thomas Weißschuh 3 months, 1 week ago

BITS_PER_LONG does not exist in UAPI headers, so can't be used by the UAPI
__GENMASK(). Instead __BITS_PER_LONG needs to be used.

When __GENMASK() was introduced in commit 3c7a8e190bc5 ("uapi: introduce uapi-friendly macros for GENMASK"),
the code was fine. A broken revert in 1e7933a575ed ("uapi: Revert "bitops: avoid integer overflow in GENMASK(_ULL)"")
introduced the incorrect usage of BITS_PER_LONG.
That was fixed in commit 11fcf368506d ("uapi: bitops: use UAPI-safe variant of BITS_PER_LONG again").
But a broken sync of the kernel headers with the tools/ headers in
commit fc92099902fb ("tools headers: Synchronize linux/bits.h with the kernel sources")
undid the fix.

Reapply the fix and while at it also fix the tools header.

Fixes: fc92099902fb ("tools headers: Synchronize linux/bits.h with the kernel sources")
Signed-off-by: Thomas Weißschuh <thomas.weissschuh@linutronix.de>
---
 include/uapi/linux/bits.h       | 4 ++--
 tools/include/uapi/linux/bits.h | 4 ++--
 2 files changed, 4 insertions(+), 4 deletions(-)

diff --git a/include/uapi/linux/bits.h b/include/uapi/linux/bits.h
index 682b406e10679dc8baa188830ab0811e7e3e13e3..a04afef9efca42f062e142fcb33f5d267512b1e5 100644
--- a/include/uapi/linux/bits.h
+++ b/include/uapi/linux/bits.h
@@ -4,9 +4,9 @@
 #ifndef _UAPI_LINUX_BITS_H
 #define _UAPI_LINUX_BITS_H
 
-#define __GENMASK(h, l) (((~_UL(0)) << (l)) & (~_UL(0) >> (BITS_PER_LONG - 1 - (h))))
+#define __GENMASK(h, l) (((~_UL(0)) << (l)) & (~_UL(0) >> (__BITS_PER_LONG - 1 - (h))))
 
-#define __GENMASK_ULL(h, l) (((~_ULL(0)) << (l)) & (~_ULL(0) >> (BITS_PER_LONG_LONG - 1 - (h))))
+#define __GENMASK_ULL(h, l) (((~_ULL(0)) << (l)) & (~_ULL(0) >> (__BITS_PER_LONG_LONG - 1 - (h))))
 
 #define __GENMASK_U128(h, l) \
 	((_BIT128((h)) << 1) - (_BIT128(l)))
diff --git a/tools/include/uapi/linux/bits.h b/tools/include/uapi/linux/bits.h
index 682b406e10679dc8baa188830ab0811e7e3e13e3..a04afef9efca42f062e142fcb33f5d267512b1e5 100644
--- a/tools/include/uapi/linux/bits.h
+++ b/tools/include/uapi/linux/bits.h
@@ -4,9 +4,9 @@
 #ifndef _UAPI_LINUX_BITS_H
 #define _UAPI_LINUX_BITS_H
 
-#define __GENMASK(h, l) (((~_UL(0)) << (l)) & (~_UL(0) >> (BITS_PER_LONG - 1 - (h))))
+#define __GENMASK(h, l) (((~_UL(0)) << (l)) & (~_UL(0) >> (__BITS_PER_LONG - 1 - (h))))
 
-#define __GENMASK_ULL(h, l) (((~_ULL(0)) << (l)) & (~_ULL(0) >> (BITS_PER_LONG_LONG - 1 - (h))))
+#define __GENMASK_ULL(h, l) (((~_ULL(0)) << (l)) & (~_ULL(0) >> (__BITS_PER_LONG_LONG - 1 - (h))))
 
 #define __GENMASK_U128(h, l) \
 	((_BIT128((h)) << 1) - (_BIT128(l)))

---
base-commit: d0b3b7b22dfa1f4b515fd3a295b3fd958f9e81af
change-id: 20250630-uapi-genmask-d36e3ffc5691

Best regards,
-- 
Thomas Weißschuh <thomas.weissschuh@linutronix.de>

Re: [PATCH] uapi: bitops: use UAPI-safe variant of BITS_PER_LONG again (2)

Posted by Yury Norov 3 months, 1 week ago

On Mon, Jun 30, 2025 at 03:02:18PM +0200, Thomas Weißschuh wrote:
> BITS_PER_LONG does not exist in UAPI headers, so can't be used by the UAPI
> __GENMASK(). Instead __BITS_PER_LONG needs to be used.
> 
> When __GENMASK() was introduced in commit 3c7a8e190bc5 ("uapi: introduce uapi-friendly macros for GENMASK"),
> the code was fine. A broken revert in 1e7933a575ed ("uapi: Revert "bitops: avoid integer overflow in GENMASK(_ULL)"")
> introduced the incorrect usage of BITS_PER_LONG.
> That was fixed in commit 11fcf368506d ("uapi: bitops: use UAPI-safe variant of BITS_PER_LONG again").
> But a broken sync of the kernel headers with the tools/ headers in
> commit fc92099902fb ("tools headers: Synchronize linux/bits.h with the kernel sources")
> undid the fix.
> 
> Reapply the fix and while at it also fix the tools header.
> 
> Fixes: fc92099902fb ("tools headers: Synchronize linux/bits.h with the kernel sources")
> Signed-off-by: Thomas Weißschuh <thomas.weissschuh@linutronix.de>

Acked-by: Yury Norov (NVIDIA) <yury.norov@gmail.com>

Arnaldo, do you want to move it yourself or with my branch?

> ---
>  include/uapi/linux/bits.h       | 4 ++--
>  tools/include/uapi/linux/bits.h | 4 ++--
>  2 files changed, 4 insertions(+), 4 deletions(-)
> 
> diff --git a/include/uapi/linux/bits.h b/include/uapi/linux/bits.h
> index 682b406e10679dc8baa188830ab0811e7e3e13e3..a04afef9efca42f062e142fcb33f5d267512b1e5 100644
> --- a/include/uapi/linux/bits.h
> +++ b/include/uapi/linux/bits.h
> @@ -4,9 +4,9 @@
>  #ifndef _UAPI_LINUX_BITS_H
>  #define _UAPI_LINUX_BITS_H
>  
> -#define __GENMASK(h, l) (((~_UL(0)) << (l)) & (~_UL(0) >> (BITS_PER_LONG - 1 - (h))))
> +#define __GENMASK(h, l) (((~_UL(0)) << (l)) & (~_UL(0) >> (__BITS_PER_LONG - 1 - (h))))
>  
> -#define __GENMASK_ULL(h, l) (((~_ULL(0)) << (l)) & (~_ULL(0) >> (BITS_PER_LONG_LONG - 1 - (h))))
> +#define __GENMASK_ULL(h, l) (((~_ULL(0)) << (l)) & (~_ULL(0) >> (__BITS_PER_LONG_LONG - 1 - (h))))
>  
>  #define __GENMASK_U128(h, l) \
>  	((_BIT128((h)) << 1) - (_BIT128(l)))
> diff --git a/tools/include/uapi/linux/bits.h b/tools/include/uapi/linux/bits.h
> index 682b406e10679dc8baa188830ab0811e7e3e13e3..a04afef9efca42f062e142fcb33f5d267512b1e5 100644
> --- a/tools/include/uapi/linux/bits.h
> +++ b/tools/include/uapi/linux/bits.h
> @@ -4,9 +4,9 @@
>  #ifndef _UAPI_LINUX_BITS_H
>  #define _UAPI_LINUX_BITS_H
>  
> -#define __GENMASK(h, l) (((~_UL(0)) << (l)) & (~_UL(0) >> (BITS_PER_LONG - 1 - (h))))
> +#define __GENMASK(h, l) (((~_UL(0)) << (l)) & (~_UL(0) >> (__BITS_PER_LONG - 1 - (h))))
>  
> -#define __GENMASK_ULL(h, l) (((~_ULL(0)) << (l)) & (~_ULL(0) >> (BITS_PER_LONG_LONG - 1 - (h))))
> +#define __GENMASK_ULL(h, l) (((~_ULL(0)) << (l)) & (~_ULL(0) >> (__BITS_PER_LONG_LONG - 1 - (h))))
>  
>  #define __GENMASK_U128(h, l) \
>  	((_BIT128((h)) << 1) - (_BIT128(l)))
> 
> ---
> base-commit: d0b3b7b22dfa1f4b515fd3a295b3fd958f9e81af
> change-id: 20250630-uapi-genmask-d36e3ffc5691
> 
> Best regards,
> -- 
> Thomas Weißschuh <thomas.weissschuh@linutronix.de>

Re: [PATCH] uapi: bitops: use UAPI-safe variant of BITS_PER_LONG again (2)

Posted by Yury Norov 3 months, 1 week ago

On Mon, Jun 30, 2025 at 10:52:54AM -0400, Yury Norov wrote:
> On Mon, Jun 30, 2025 at 03:02:18PM +0200, Thomas Weißschuh wrote:
> > BITS_PER_LONG does not exist in UAPI headers, so can't be used by the UAPI
> > __GENMASK(). Instead __BITS_PER_LONG needs to be used.
> > 
> > When __GENMASK() was introduced in commit 3c7a8e190bc5 ("uapi: introduce uapi-friendly macros for GENMASK"),
> > the code was fine. A broken revert in 1e7933a575ed ("uapi: Revert "bitops: avoid integer overflow in GENMASK(_ULL)"")
> > introduced the incorrect usage of BITS_PER_LONG.
> > That was fixed in commit 11fcf368506d ("uapi: bitops: use UAPI-safe variant of BITS_PER_LONG again").
> > But a broken sync of the kernel headers with the tools/ headers in
> > commit fc92099902fb ("tools headers: Synchronize linux/bits.h with the kernel sources")
> > undid the fix.
> > 
> > Reapply the fix and while at it also fix the tools header.
> > 
> > Fixes: fc92099902fb ("tools headers: Synchronize linux/bits.h with the kernel sources")
> > Signed-off-by: Thomas Weißschuh <thomas.weissschuh@linutronix.de>
> 
> Acked-by: Yury Norov (NVIDIA) <yury.norov@gmail.com>
> 
> Arnaldo, do you want to move it yourself or with my branch?

OK, added this in bitmap-for-next together with the MAINTAINERS patch.

Thanks Tomas for looking after that!

Thanks,
Yury

Re: [PATCH] uapi: bitops: use UAPI-safe variant of BITS_PER_LONG again (2)

Posted by Thomas Weißschuh 3 months ago

On Tue, Jul 01, 2025 at 05:37:40PM -0400, Yury Norov wrote:
> On Mon, Jun 30, 2025 at 10:52:54AM -0400, Yury Norov wrote:
> > On Mon, Jun 30, 2025 at 03:02:18PM +0200, Thomas Weißschuh wrote:
> > > BITS_PER_LONG does not exist in UAPI headers, so can't be used by the UAPI
> > > __GENMASK(). Instead __BITS_PER_LONG needs to be used.
> > > 
> > > When __GENMASK() was introduced in commit 3c7a8e190bc5 ("uapi: introduce uapi-friendly macros for GENMASK"),
> > > the code was fine. A broken revert in 1e7933a575ed ("uapi: Revert "bitops: avoid integer overflow in GENMASK(_ULL)"")
> > > introduced the incorrect usage of BITS_PER_LONG.
> > > That was fixed in commit 11fcf368506d ("uapi: bitops: use UAPI-safe variant of BITS_PER_LONG again").
> > > But a broken sync of the kernel headers with the tools/ headers in
> > > commit fc92099902fb ("tools headers: Synchronize linux/bits.h with the kernel sources")
> > > undid the fix.
> > > 
> > > Reapply the fix and while at it also fix the tools header.
> > > 
> > > Fixes: fc92099902fb ("tools headers: Synchronize linux/bits.h with the kernel sources")
> > > Signed-off-by: Thomas Weißschuh <thomas.weissschuh@linutronix.de>
> > 
> > Acked-by: Yury Norov (NVIDIA) <yury.norov@gmail.com>
> > 
> > Arnaldo, do you want to move it yourself or with my branch?
> 
> OK, added this in bitmap-for-next together with the MAINTAINERS patch.

Thanks!

Any chance to get the fix into v6.16 again?
This currently triggers warnings in my code in -next [0], which is intended
to go into v6.17.

[0] https://lore.kernel.org/all/20250708160830.36ddf20f@canb.auug.org.au/


Thomas

Re: [PATCH] uapi: bitops: use UAPI-safe variant of BITS_PER_LONG again (2)

Posted by Yury Norov 3 months ago

On Tue, Jul 08, 2025 at 08:26:49AM +0200, Thomas Weißschuh wrote:
> On Tue, Jul 01, 2025 at 05:37:40PM -0400, Yury Norov wrote:
> > On Mon, Jun 30, 2025 at 10:52:54AM -0400, Yury Norov wrote:
> > > On Mon, Jun 30, 2025 at 03:02:18PM +0200, Thomas Weißschuh wrote:
> > > > BITS_PER_LONG does not exist in UAPI headers, so can't be used by the UAPI
> > > > __GENMASK(). Instead __BITS_PER_LONG needs to be used.
> > > > 
> > > > When __GENMASK() was introduced in commit 3c7a8e190bc5 ("uapi: introduce uapi-friendly macros for GENMASK"),
> > > > the code was fine. A broken revert in 1e7933a575ed ("uapi: Revert "bitops: avoid integer overflow in GENMASK(_ULL)"")
> > > > introduced the incorrect usage of BITS_PER_LONG.
> > > > That was fixed in commit 11fcf368506d ("uapi: bitops: use UAPI-safe variant of BITS_PER_LONG again").
> > > > But a broken sync of the kernel headers with the tools/ headers in
> > > > commit fc92099902fb ("tools headers: Synchronize linux/bits.h with the kernel sources")
> > > > undid the fix.
> > > > 
> > > > Reapply the fix and while at it also fix the tools header.
> > > > 
> > > > Fixes: fc92099902fb ("tools headers: Synchronize linux/bits.h with the kernel sources")
> > > > Signed-off-by: Thomas Weißschuh <thomas.weissschuh@linutronix.de>
> > > 
> > > Acked-by: Yury Norov (NVIDIA) <yury.norov@gmail.com>
> > > 
> > > Arnaldo, do you want to move it yourself or with my branch?
> > 
> > OK, added this in bitmap-for-next together with the MAINTAINERS patch.
> 
> Thanks!
> 
> Any chance to get the fix into v6.16 again?

Yes, just sent a pull request.

> This currently triggers warnings in my code in -next [0], which is intended
> to go into v6.17.
> 
> [0] https://lore.kernel.org/all/20250708160830.36ddf20f@canb.auug.org.au/
> 
> 
> Thomas