linux-kernel - Re: [PATCH v2] arch_topology: Make cluster topology span at least SMT CPUs

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite: Windows password security audit tool. GUI, reports in PDF.

[<prev] [next>] [<thread-prev] [day] [month] [year] [list]

Message-ID: <a76d1153-dc5a-d26c-75d2-4fdb48a9641d@huawei.com>
Date:   Mon, 5 Sep 2022 11:41:38 +0800
From:   Yicong Yang <yangyicong@...wei.com>
To:     Greg KH <gregkh@...uxfoundation.org>
CC:     <yangyicong@...ilicon.com>, <sudeep.holla@....com>,
        <vincent.guittot@...aro.org>, <ionela.voinescu@....com>,
        <linux-kernel@...r.kernel.org>, <rafael@...nel.org>,
        <21cnbao@...il.com>, <jonathan.cameron@...wei.com>,
        <linuxarm@...wei.com>, <prime.zeng@...wei.com>
Subject: Re: [PATCH v2] arch_topology: Make cluster topology span at least SMT
 CPUs

On 2022/9/3 14:00, Greg KH wrote:
> On Thu, Aug 25, 2022 at 05:20:07PM +0800, Yicong Yang wrote:
>> From: Yicong Yang <yangyicong@...ilicon.com>
>>
>> Currently cpu_clustergroup_mask() will return CPU mask if cluster span more
>> or the same CPUs as cpu_coregroup_mask(). This will result topology borken
>> on non-Cluster SMT machines when building with CONFIG_SCHED_CLUSTER=y.
>>
>> Test with:
>> qemu-system-aarch64 -enable-kvm -machine virt \
>>  -net none \
>>  -cpu host \
>>  -bios ./QEMU_EFI.fd \
>>  -m 2G \
>>  -smp 48,sockets=2,cores=12,threads=2 \
>>  -kernel $Image \
>>  -initrd $Rootfs \
>>  -nographic \
>>  -append "rdinit=init console=ttyAMA0 sched_verbose loglevel=8"
>>
>> We'll get below error:
>> [    3.084568] BUG: arch topology borken
>> [    3.084570]      the SMT domain not a subset of the CLS domain
>>
>> Since cluster is a level higher than SMT, fix this by making cluster
>> spans at least SMT CPUs.
>>
>> Cc: Sudeep Holla <sudeep.holla@....com>
>> Cc: Vincent Guittot <vincent.guittot@...aro.org>
>> Cc: Ionela Voinescu <ionela.voinescu@....com>
>> Fixes: bfcc4397435d ("arch_topology: Limit span of cpu_clustergroup_mask()")
>> Signed-off-by: Yicong Yang <yangyicong@...ilicon.com>
>> Reviewed-by: Sudeep Holla <sudeep.holla@....com>
>> Reviewed-by: Ionela Voinescu <ionela.voinescu@....com>
>> ---
>> Change since v1:
>> - mention the kernel config CONFIG_SCHED_CLUSTER=y, per Ionela
>> Link:https://lore.kernel.org/lkml/20220823073044.58697-1-yangyicong@huawei.com/
>>
>>  drivers/base/arch_topology.c | 2 +-
>>  1 file changed, 1 insertion(+), 1 deletion(-)
>>
>> diff --git a/drivers/base/arch_topology.c b/drivers/base/arch_topology.c
>> index 0424b59b695e..0056a1273275 100644
>> --- a/drivers/base/arch_topology.c
>> +++ b/drivers/base/arch_topology.c
>> @@ -724,7 +724,7 @@ const struct cpumask *cpu_clustergroup_mask(int cpu)
>>  	 */
>>  	if (cpumask_subset(cpu_coregroup_mask(cpu),
>>  			   &cpu_topology[cpu].cluster_sibling))
>> -		return get_cpu_mask(cpu);
>> +		return cpu_smt_mask(cpu);
> 
> As the kernel-test robot reported, this breaks the build on some
> architectures, so I'll go and revert it from my tree now.
> 

Sorry for the regression.

The build error is because cpu_smt_mask() is defined only when CONFIG_SCHED_SMT=y
so we'll fail when CONFIG_SCHED_SMT=n. Should use topology_sibling_cpumask() to
get SMT mask here. Will respin this.

Thanks.