[PATCH] x86/cacheinfo: fix a -Wtype-limits warning

[PATCH] x86/cacheinfo: fix a -Wtype-limits warning

[Qian Cai]

cpuinfo_x86.x86_model is an unsigned type, so compares it against zero
will generate a compilation warning,

arch/x86/kernel/cpu/cacheinfo.c: In function
'cacheinfo_amd_init_llc_id':
arch/x86/kernel/cpu/cacheinfo.c:662:19: warning: comparison is always
true due to limited range of data type [-Wtype-limits]

Signed-off-by: Qian Cai <cai@lca.pw>
---
 arch/x86/kernel/cpu/cacheinfo.c | 3 +--
 1 file changed, 1 insertion(+), 2 deletions(-)

diff --git a/arch/x86/kernel/cpu/cacheinfo.c b/arch/x86/kernel/cpu/cacheinfo.c
index 395d46f78582..c7503be92f35 100644
--- a/arch/x86/kernel/cpu/cacheinfo.c
+++ b/arch/x86/kernel/cpu/cacheinfo.c
@@ -658,8 +658,7 @@ void cacheinfo_amd_init_llc_id(struct cpuinfo_x86 *c, int cpu, u8 node_id)
 	if (c->x86 < 0x17) {
 		/* LLC is at the node level. */
 		per_cpu(cpu_llc_id, cpu) = node_id;
-	} else if (c->x86 == 0x17 &&
-		   c->x86_model >= 0 && c->x86_model <= 0x1F) {
+	} else if (c->x86 == 0x17 && c->x86_model <= 0x1F) {
 		/*
 		 * LLC is at the core complex level.
 		 * Core complex ID is ApicId[3] for these processors.
-- 
1.8.3.1

Re: [PATCH] x86/cacheinfo: fix a -Wtype-limits warning

[Sean Christopherson]

On Wed, Jun 05, 2019 at 03:40:54PM -0400, Qian Cai wrote:
> cpuinfo_x86.x86_model is an unsigned type, so compares it against zero
> will generate a compilation warning,
> 
> arch/x86/kernel/cpu/cacheinfo.c: In function
> 'cacheinfo_amd_init_llc_id':
> arch/x86/kernel/cpu/cacheinfo.c:662:19: warning: comparison is always
> true due to limited range of data type [-Wtype-limits]
> 
> Signed-off-by: Qian Cai <cai@lca.pw>
> ---
>  arch/x86/kernel/cpu/cacheinfo.c | 3 +--
>  1 file changed, 1 insertion(+), 2 deletions(-)
> 
> diff --git a/arch/x86/kernel/cpu/cacheinfo.c b/arch/x86/kernel/cpu/cacheinfo.c
> index 395d46f78582..c7503be92f35 100644
> --- a/arch/x86/kernel/cpu/cacheinfo.c
> +++ b/arch/x86/kernel/cpu/cacheinfo.c
> @@ -658,8 +658,7 @@ void cacheinfo_amd_init_llc_id(struct cpuinfo_x86 *c, int cpu, u8 node_id)
>  	if (c->x86 < 0x17) {
>  		/* LLC is at the node level. */
>  		per_cpu(cpu_llc_id, cpu) = node_id;
> -	} else if (c->x86 == 0x17 &&
> -		   c->x86_model >= 0 && c->x86_model <= 0x1F) {
> +	} else if (c->x86 == 0x17 && c->x86_model <= 0x1F) {

Might be worth calling out in the changelog that 'c->x86 == 0x17' is true
if and only if c->x86_model was explicitly set by cpu_detect(), i.e. the
patch is correct even if the original intent was a misguided attempt to
check that x86_model has been set.

Fixes: 68091ee7ac3c ("x86/CPU/AMD: Calculate last level cache ID from number of sharing threads")
Reviewed-by: Sean Christopherson <sean.j.christopherson@intel.com>

>  		/*
>  		 * LLC is at the core complex level.
>  		 * Core complex ID is ApicId[3] for these processors.
> -- 
> 1.8.3.1
> 

Re: [PATCH] x86/cacheinfo: fix a -Wtype-limits warning

[Borislav Petkov]

On Wed, Jun 05, 2019 at 01:07:04PM -0700, Sean Christopherson wrote:
> Might be worth calling out in the changelog that 'c->x86 == 0x17' is true
> if and only if c->x86_model was explicitly set by cpu_detect(), i.e. the
> patch is correct even if the original intent was a misguided attempt to
> check that x86_model has been set.

Are you thinking about some sick virt scenario where base CPUID level is < 1?

In this particular case, there's a guard at the beginning of
cacheinfo_amd_init_llc_id():

        if (!cpuid_edx(0x80000006))
                return;

but if there's CPUs which have CPUID 0x80000006 but base CPUID level is
< 1, then that's their problem.

-- 
Regards/Gruss,
    Boris.

Good mailing practices for 400: avoid top-posting and trim the reply.

Re: [PATCH] x86/cacheinfo: fix a -Wtype-limits warning

[Sean Christopherson]

On Wed, Jun 19, 2019 at 07:01:27PM +0200, Borislav Petkov wrote:
> On Wed, Jun 05, 2019 at 01:07:04PM -0700, Sean Christopherson wrote:
> > Might be worth calling out in the changelog that 'c->x86 == 0x17' is true
> > if and only if c->x86_model was explicitly set by cpu_detect(), i.e. the
> > patch is correct even if the original intent was a misguided attempt to
> > check that x86_model has been set.
> 
> Are you thinking about some sick virt scenario where base CPUID level is < 1?

Ha, no.  My comment was that it'd be worth explaining that the original
'c->x86_model >= 0' check was completely bogus, even if the intent was
something like 'c->x86_model != 0'.

> In this particular case, there's a guard at the beginning of
> cacheinfo_amd_init_llc_id():
> 
>         if (!cpuid_edx(0x80000006))
>                 return;
> 
> but if there's CPUs which have CPUID 0x80000006 but base CPUID level is
> < 1, then that's their problem.
> 
> -- 
> Regards/Gruss,
>     Boris.
> 
> Good mailing practices for 400: avoid top-posting and trim the reply.

Re: [PATCH] x86/cacheinfo: fix a -Wtype-limits warning

[Borislav Petkov]

On Wed, Jun 19, 2019 at 10:12:49AM -0700, Sean Christopherson wrote:
> Ha, no.  My comment was that it'd be worth explaining that the original
> 'c->x86_model >= 0' check was completely bogus, even if the intent was
> something like 'c->x86_model != 0'.

Nah, the intent was, I believe, to convert a model range into a
conditional. If ->x86_model is not set we have bigger problems.

-- 
Regards/Gruss,
    Boris.

Good mailing practices for 400: avoid top-posting and trim the reply.