[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Message-ID: <20241213162443.GGZ1xfyw_EZBrn1i4B@fat_crate.local>
Date: Fri, 13 Dec 2024 17:24:43 +0100
From: Borislav Petkov <bp@...en8.de>
To: Dave Hansen <dave.hansen@...ux.intel.com>
Cc: linux-kernel@...r.kernel.org, x86@...nel.org, tglx@...utronix.de,
kan.liang@...ux.intel.com, mingo@...nel.org, peterz@...radead.org,
tony.luck@...el.com, pawan.kumar.gupta@...ux.intel.com
Subject: Re: [PATCH 2/5] x86/cpu: Expose only stepping min/max interface
On Fri, Dec 06, 2024 at 11:38:32AM -0800, Dave Hansen wrote:
>
> From: Dave Hansen <dave.hansen@...ux.intel.com>
>
> The x86_match_cpu() infrastructure can match CPU steppings. Since
> there are only 16 possible steppings, the matching infrastructure goes
> all out and stores the stepping match as a bitmap. That means it can
> match any possible steppings in a single list entry. Fun.
>
> But it exposes this bitmap to each of the X86_MATCH_*() helpers when
> none of them really need a bitmap. It makes up for this by exporting a
> helper (X86_STEPPINGS()) which converts a contiguous stepping range
> into the bitmap which every single user leverages.
>
> Instead of a bitmap, have the main helper for this sort of thing
> (X86_MATCH_VFM_STEPPING()) just take a stepping range. This ends up
> actually being even more compact than before.
Yap, better.
> Leave the helper in place (renamed to __X86_STEPPINGS()) to make it
> more clear what is going on instead of just having a random GENMASK()
> in the middle of an already complicated macro.
>
> One oddity that I hit was this macro:
>
> #define VULNBL_INTEL_STEPPINGS(vfm, max_stepping, issues) \
> X86_MATCH_VFM_STEPPINGS(vfm, X86_STEPPING_MIN, max_stepping, issues)
>
> It *could* have been converted over to take a min/max stepping value
> for each entry. But that would have been a bit too verbose and would
> prevent the one oddball in the list (INTEL_COMETLAKE_L stepping 0)
> from sticking out.
>
> Instead, just have it take a *maximum* stepping and imply that the match
> is from 0=>max_stepping. This is functional for all the cases now and
> also retains the nice property of having INTEL_COMETLAKE_L stepping 0
> stick out like a sore thumb.
>
> Signed-off-by: Dave Hansen <dave.hansen@...ux.intel.com>
> Suggested-by: Ingo Molnar <mingo@...nel.org>
> ---
>
> b/arch/x86/include/asm/cpu_device_id.h | 15 +++---
> b/arch/x86/kernel/apic/apic.c | 18 +++----
> b/arch/x86/kernel/cpu/common.c | 78 ++++++++++++++++-----------------
> b/drivers/edac/i10nm_base.c | 20 ++++----
> b/drivers/edac/skx_base.c | 2
> b/include/linux/mod_devicetable.h | 2
> 6 files changed, 69 insertions(+), 66 deletions(-)
You missed a spot:
drivers/edac/i10nm_base.c:951:90: error: macro "X86_MATCH_VFM_STEPPINGS" requires 4 arguments, but only 3 given
951 | X86_MATCH_VFM_STEPPINGS(INTEL_ATOM_DARKMONT_X, X86_STEPPINGS(0x0, 0xf), &gnr_cfg),
| ^
In file included from drivers/edac/i10nm_base.c:10:
./arch/x86/include/asm/cpu_device_id.h:221: note: macro "X86_MATCH_VFM_STEPPINGS" defined here
221 | #define X86_MATCH_VFM_STEPPINGS(vfm, min_step, max_step, data) \
|
drivers/edac/i10nm_base.c:951:9: error: ‘X86_MATCH_VFM_STEPPINGS’ undeclared here (not in a function)
951 | X86_MATCH_VFM_STEPPINGS(INTEL_ATOM_DARKMONT_X, X86_STEPPINGS(0x0, 0xf), &gnr_cfg),
| ^~~~~~~~~~~~~~~~~~~~~~~
make[4]: *** [scripts/Makefile.build:194: drivers/edac/i10nm_base.o] Error 1
make[4]: *** Waiting for unfinished jobs....
make[3]: *** [scripts/Makefile.build:440: drivers/edac] Error 2
make[2]: *** [scripts/Makefile.build:440: drivers] Error 2
make[2]: *** Waiting for unfinished jobs....
make[1]: *** [/mnt/kernel/kernel/2nd/linux/Makefile:1989: .] Error 2
make: *** [Makefile:251: __sub-make] Error 2
> diff -puN drivers/edac/i10nm_base.c~zap-X86_STEPPINGS drivers/edac/i10nm_base.c
> --- a/drivers/edac/i10nm_base.c~zap-X86_STEPPINGS 2024-12-06 11:33:16.187148838 -0800
> +++ b/drivers/edac/i10nm_base.c 2024-12-06 11:33:16.191148995 -0800
> @@ -938,16 +938,16 @@ static struct res_config gnr_cfg = {
> };
>
> static const struct x86_cpu_id i10nm_cpuids[] = {
> - X86_MATCH_VFM_STEPPINGS(INTEL_ATOM_TREMONT_D, X86_STEPPINGS(0x0, 0x3), &i10nm_cfg0),
> - X86_MATCH_VFM_STEPPINGS(INTEL_ATOM_TREMONT_D, X86_STEPPINGS(0x4, 0xf), &i10nm_cfg1),
> - X86_MATCH_VFM_STEPPINGS(INTEL_ICELAKE_X, X86_STEPPINGS(0x0, 0x3), &i10nm_cfg0),
> - X86_MATCH_VFM_STEPPINGS(INTEL_ICELAKE_X, X86_STEPPINGS(0x4, 0xf), &i10nm_cfg1),
> - X86_MATCH_VFM_STEPPINGS(INTEL_ICELAKE_D, X86_STEPPINGS(0x0, 0xf), &i10nm_cfg1),
> - X86_MATCH_VFM_STEPPINGS(INTEL_SAPPHIRERAPIDS_X, X86_STEPPINGS(0x0, 0xf), &spr_cfg),
> - X86_MATCH_VFM_STEPPINGS(INTEL_EMERALDRAPIDS_X, X86_STEPPINGS(0x0, 0xf), &spr_cfg),
> - X86_MATCH_VFM_STEPPINGS(INTEL_GRANITERAPIDS_X, X86_STEPPINGS(0x0, 0xf), &gnr_cfg),
> - X86_MATCH_VFM_STEPPINGS(INTEL_ATOM_CRESTMONT_X, X86_STEPPINGS(0x0, 0xf), &gnr_cfg),
> - X86_MATCH_VFM_STEPPINGS(INTEL_ATOM_CRESTMONT, X86_STEPPINGS(0x0, 0xf), &gnr_cfg),
> + X86_MATCH_VFM_STEPPINGS(INTEL_ATOM_TREMONT_D, 0x0, 0x3, &i10nm_cfg0),
> + X86_MATCH_VFM_STEPPINGS(INTEL_ATOM_TREMONT_D, 0x4, 0xf, &i10nm_cfg1),
> + X86_MATCH_VFM_STEPPINGS(INTEL_ICELAKE_X, 0x0, 0x3, &i10nm_cfg0),
> + X86_MATCH_VFM_STEPPINGS(INTEL_ICELAKE_X, 0x4, 0xf, &i10nm_cfg1),
> + X86_MATCH_VFM_STEPPINGS(INTEL_ICELAKE_D, 0x0, 0xf, &i10nm_cfg1),
> + X86_MATCH_VFM_STEPPINGS(INTEL_SAPPHIRERAPIDS_X, 0x0, 0xf, &spr_cfg),
> + X86_MATCH_VFM_STEPPINGS(INTEL_EMERALDRAPIDS_X, 0x0, 0xf, &spr_cfg),
> + X86_MATCH_VFM_STEPPINGS(INTEL_GRANITERAPIDS_X, 0x0, 0xf, &gnr_cfg),
> + X86_MATCH_VFM_STEPPINGS(INTEL_ATOM_CRESTMONT_X, 0x0, 0xf, &gnr_cfg),
> + X86_MATCH_VFM_STEPPINGS(INTEL_ATOM_CRESTMONT, 0x0, 0xf, &gnr_cfg),
Aren't those supposed to be:
X86_MATCH_VFM_STEPPINGS(INTEL_ATOM_CRESTMONT, X86_STEPPING_MIN, X86_STEPPING_MAX, &gnr_cfg),
And while we're adding new defines, can we shorten them too?
X86_MATCH_VFM_STP(INTEL_ATOM_CRESTMONT, X86_STP_MIN, X86_STP_MAX, &gnr_cfg),
all that "STEPPING" is screaming at me! :-P
--
Regards/Gruss,
Boris.
https://people.kernel.org/tglx/notes-about-netiquette
Powered by blists - more mailing lists