Re: [linux-next:master] BUILD REGRESSION a5444cde9dc2120612e50fc5a56c975e67a041fb

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

 



On Mon, Mar 19, 2018 at 06:10:13PM -0700, Randy Dunlap wrote:
> On 03/19/2018 04:33 PM, kbuild test robot wrote:
> > tree/branch: https://git.kernel.org/pub/scm/linux/kernel/git/next/linux-next.git  master
> > branch HEAD: a5444cde9dc2120612e50fc5a56c975e67a041fb  Add linux-next specific files for 20180319
> > 
> > Regressions in current branch:
> > 
> > ERROR: "__sw_hweight8" [drivers/net/wireless/mediatek/mt76/mt76.ko] undefined!
> 
> Well, the driver could do:
> 
> drivers/net/wireless/mediatek/mt76/Kconfig:
> 
> +	select GENERIC_HWEIGHT
> 
> but maybe arch/ix64/Kconfig (where the build error is) could be asked to do:
> 
> config GENERIC_HWEIGHT
> 	def_bool y
> 
> like 23 other $arch-es do.  Aha, ia64 provides inline functions via some
> a twisty maze of header files.
> 
> Tony, Fengguang, what header(s) should be used to reach __arch_hweight8()?

Looks like a few architectures have their own __arch_hweight8
(alpha, blackfin, ia64, mips, powerpc, sparc, tile and x86)

everyone except x86 puts them in <asm/bitops.h> ... x86 has a
<asm/arch_hweight.h>

Likely that the best solution would be to match how x86 does
this and move the hweight defines out of bitops.h.  But it all
seems very messy :-(

-Tony




[Index of Archives]     [Linux ARM Kernel]     [Linux ARM]     [Linux Omap]     [Fedora ARM]     [IETF Annouce]     [Bugtraq]     [Linux OMAP]     [Linux MIPS]     [eCos]     [Asterisk Internet PBX]     [Linux API]

  Powered by Linux