On Thu, Aug 22, 2019 at 12:23:06PM +0200, Peter Zijlstra wrote:
> Lots of variation has crept in; time to collapse the lot again.
Conceptually good. But I applied the series on top of tip/master
and got a build error:
CC arch/x86/kernel/cpu/common.o
arch/x86/kernel/cpu/common.c:1031:19: error: ‘INTEL_FAM6_ATOM_SILVERMONT_X’
undeclared here (not in a function); did you mean
‘INTEL_FAM6_ATOM_SILVERMONT_D’?
VULNWL(INTEL, 6, INTEL_FAM6_##model, whitelist)
^~~~~~~~~~~
arch/x86/kernel/cpu/common.c:1028:35: note: in definition of macro ‘VULNWL’
{ X86_VENDOR_##_vendor, _family, _model, X86_FEATURE_ANY, _whitelist }
^~~~~~
arch/x86/kernel/cpu/common.c:1053:2: note: in expansion of macro ‘VULNWL_INTEL’
VULNWL_INTEL(ATOM_SILVERMONT_X, NO_SSB | NO_L1TF | MSBDS_ONLY | NO_SWAPGS),
^~~~~~~~~~~~
arch/x86/kernel/cpu/common.c:1031:19: error: ‘INTEL_FAM6_ATOM_GOLDMONT_X’
undeclared here (not in a function); did you mean ‘INTEL_FAM6_ATOM_GOLDMONT_D’?
VULNWL(INTEL, 6, INTEL_FAM6_##model, whitelist)
^~~~~~~~~~~
arch/x86/kernel/cpu/common.c:1028:35: note: in definition of macro ‘VULNWL’
{ X86_VENDOR_##_vendor, _family, _model, X86_FEATURE_ANY, _whitelist }
^~~~~~
arch/x86/kernel/cpu/common.c:1064:2: note: in expansion of macro ‘VULNWL_INTEL’
VULNWL_INTEL(ATOM_GOLDMONT_X, NO_MDS | NO_L1TF | NO_SWAPGS),
^~~~~~~~~~~~
make[3]: *** [scripts/Makefile.build:281: arch/x86/kernel/cpu/common.o] Error 1
Looks like your scripts didn't anticipate the CPP gymnastics like:
#define VULNWL_INTEL(model, whitelist) \
VULNWL(INTEL, 6, INTEL_FAM6_##model, whitelist)
-Tony
