linux-parisc.vger.kernel.org archive mirror
 help / color / mirror / Atom feed
* Re: drivers/scsi/ncr53c8xx.c:5306:9: sparse: sparse: cast truncates bits from constant value (58f becomes 8f)
       [not found] <202005160227.h6Ieqnmz%lkp@intel.com>
@ 2020-05-15 19:00 ` Matthew Wilcox
  2020-05-18  0:34   ` [kbuild-all] " Philip Li
  0 siblings, 1 reply; 2+ messages in thread
From: Matthew Wilcox @ 2020-05-15 19:00 UTC (permalink / raw)
  To: kbuild test robot
  Cc: Michal Simek, kbuild-all, linux-kernel, Andrew Morton,
	Linux Memory Management List, Stefan Asserhall, linux-scsi,
	linux-parisc

On Sat, May 16, 2020 at 02:20:38AM +0800, kbuild test robot wrote:
> tree:   https://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git master
> head:   051e6b7e34b9bd24f46725f74994a4d3a653966e
> commit: 06e85c7e9a1c1356038936566fc23f7c0d363b96 asm-generic: fix unistd_32.h generation format
> date:   5 weeks ago

I don't see how that commit in any way reflects this error message.

> reproduce:
>         # apt-get install sparse
>         # sparse version: v0.6.1-193-gb8fad4bc-dirty
>         git checkout 06e85c7e9a1c1356038936566fc23f7c0d363b96
>         make ARCH=x86_64 allmodconfig

I can't even see a way to build the ncr53c8xx module with this config.
Unless somebody reenabled EISA on x86, the only way I can see to
still build this driver is on PA-RISC with the ZALON code.

>         make C=1 CF='-fdiagnostic-prefix -D__CHECK_ENDIAN__'
> 
> If you fix the issue, kindly add following tag as appropriate
> Reported-by: kbuild test robot <lkp@intel.com>
> 
> 
> sparse warnings: (new ones prefixed by >>)
> 
> >> drivers/scsi/ncr53c8xx.c:5306:9: sparse: sparse: cast truncates bits from constant value (58f becomes 8f)
> 
> ^1da177e4c3f41 Linus Torvalds 2005-04-16 @5306  	OUTW (nc_sien , STO|HTH|MA|SGE|UDC|RST|PAR);

This seems entirely intentional.

Something like this should do the job (whitespace damaged):

+++ b/drivers/scsi/ncr53c8xx.h
@@ -407,7 +407,7 @@
 
 #ifdef CONFIG_SCSI_NCR53C8XX_NO_WORD_TRANSFERS
 /* Only 8 or 32 bit transfers allowed */
-#define OUTW_OFF(o, val)       do { writeb((char)((val) >> 8), (char __iomem *)np->reg + ncr_offw(o)); writeb((char)(val), (char __iomem *)np->reg + ncr_offw(o) + 1); } while (0)
+#define OUTW_OFF(o, val)       do { writeb((char)((val) >> 8), (char __iomem *)np->reg + ncr_offw(o)); writeb((char)((val) & 0xff), (char __iomem *)np->reg + ncr_offw(o) + 1); } while (0)
 #else
 #define OUTW_OFF(o, val)       writew_raw((val), (char __iomem *)np->reg + ncr_offw(o))
 #endif



^ permalink raw reply	[flat|nested] 2+ messages in thread

* Re: [kbuild-all] Re: drivers/scsi/ncr53c8xx.c:5306:9: sparse: sparse: cast truncates bits from constant value (58f becomes 8f)
  2020-05-15 19:00 ` drivers/scsi/ncr53c8xx.c:5306:9: sparse: sparse: cast truncates bits from constant value (58f becomes 8f) Matthew Wilcox
@ 2020-05-18  0:34   ` Philip Li
  0 siblings, 0 replies; 2+ messages in thread
From: Philip Li @ 2020-05-18  0:34 UTC (permalink / raw)
  To: Matthew Wilcox
  Cc: kbuild test robot, Michal Simek, kbuild-all, linux-kernel,
	Andrew Morton, Linux Memory Management List, Stefan Asserhall,
	linux-scsi, linux-parisc

On Fri, May 15, 2020 at 12:00:26PM -0700, Matthew Wilcox wrote:
> On Sat, May 16, 2020 at 02:20:38AM +0800, kbuild test robot wrote:
> > tree:   https://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git master
> > head:   051e6b7e34b9bd24f46725f74994a4d3a653966e
> > commit: 06e85c7e9a1c1356038936566fc23f7c0d363b96 asm-generic: fix unistd_32.h generation format
> > date:   5 weeks ago
> 
> I don't see how that commit in any way reflects this error message.
> 
> > reproduce:
> >         # apt-get install sparse
> >         # sparse version: v0.6.1-193-gb8fad4bc-dirty
> >         git checkout 06e85c7e9a1c1356038936566fc23f7c0d363b96
> >         make ARCH=x86_64 allmodconfig
> 
> I can't even see a way to build the ncr53c8xx module with this config.
> Unless somebody reenabled EISA on x86, the only way I can see to
> still build this driver is on PA-RISC with the ZALON code.
sorry, the reproduce steps here is wrong, it is not for x86_64. We will
fix this.

> 
> >         make C=1 CF='-fdiagnostic-prefix -D__CHECK_ENDIAN__'
> > 
> > If you fix the issue, kindly add following tag as appropriate
> > Reported-by: kbuild test robot <lkp@intel.com>
> > 
> > 
> > sparse warnings: (new ones prefixed by >>)
> > 
> > >> drivers/scsi/ncr53c8xx.c:5306:9: sparse: sparse: cast truncates bits from constant value (58f becomes 8f)
> > 
> > ^1da177e4c3f41 Linus Torvalds 2005-04-16 @5306  	OUTW (nc_sien , STO|HTH|MA|SGE|UDC|RST|PAR);
> 
> This seems entirely intentional.
> 
> Something like this should do the job (whitespace damaged):
> 
> +++ b/drivers/scsi/ncr53c8xx.h
> @@ -407,7 +407,7 @@
>  
>  #ifdef CONFIG_SCSI_NCR53C8XX_NO_WORD_TRANSFERS
>  /* Only 8 or 32 bit transfers allowed */
> -#define OUTW_OFF(o, val)       do { writeb((char)((val) >> 8), (char __iomem *)np->reg + ncr_offw(o)); writeb((char)(val), (char __iomem *)np->reg + ncr_offw(o) + 1); } while (0)
> +#define OUTW_OFF(o, val)       do { writeb((char)((val) >> 8), (char __iomem *)np->reg + ncr_offw(o)); writeb((char)((val) & 0xff), (char __iomem *)np->reg + ncr_offw(o) + 1); } while (0)
>  #else
>  #define OUTW_OFF(o, val)       writew_raw((val), (char __iomem *)np->reg + ncr_offw(o))
>  #endif
> 
> _______________________________________________
> kbuild-all mailing list -- kbuild-all@lists.01.org
> To unsubscribe send an email to kbuild-all-leave@lists.01.org

^ permalink raw reply	[flat|nested] 2+ messages in thread

end of thread, other threads:[~2020-05-18  0:35 UTC | newest]

Thread overview: 2+ messages (download: mbox.gz / follow: Atom feed)
-- links below jump to the message on this page --
     [not found] <202005160227.h6Ieqnmz%lkp@intel.com>
2020-05-15 19:00 ` drivers/scsi/ncr53c8xx.c:5306:9: sparse: sparse: cast truncates bits from constant value (58f becomes 8f) Matthew Wilcox
2020-05-18  0:34   ` [kbuild-all] " Philip Li

This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).