[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Message-ID: <7cca5e74-6626-1c8b-9309-47b9f5d4395f@arm.com>
Date: Thu, 23 Mar 2023 18:58:53 +0100
From: Pierre Gondois <pierre.gondois@....com>
To: Yicong Yang <yangyicong@...wei.com>, gregkh@...uxfoundation.org,
rafael@...nel.org, sudeep.holla@....com, palmer@...osinc.com,
linux-kernel@...r.kernel.org
Cc: prime.zeng@...ilicon.com, linuxarm@...wei.com,
yangyicong@...ilicon.com
Subject: Re: [PATCH] cacheinfo: Fix LLC is not exported through sysfs
Hello Yicong,
FWIW, I think the patch is correct and I could reproduce the issue.
On 3/23/23 13:25, Yicong Yang wrote:
> From: Yicong Yang <yangyicong@...ilicon.com>
>
> After entering 6.3-rc1 the LLC cacheinfo is not exported on our ACPI
> based arm64 server. This is because the LLC cacheinfo is partly reset
> when secondary CPUs boot up. On arm64 the primary cpu will allocate
> and setup cacheinfo:
> init_cpu_topology()
> for_each_possible_cpu()
> fetch_cache_info() // Allocate cacheinfo and init levels
> detect_cache_attributes()
> cache_shared_cpu_map_setup()
> if (!last_level_cache_is_valid()) // not valid, setup LLC
> cache_setup_properties() // setup LLC
>
> On secondary CPU boot up:
> detect_cache_attributes()
> populate_cache_leaves()
> get_cache_type() // Get cache type from clidr_el1,
> // for LLC type=CACHE_TYPE_NOCACHE
> cache_shared_cpu_map_setup()
> if (!last_level_cache_is_valid()) // Valid and won't go to this branch,
> // leave LLC's type=CACHE_TYPE_NOCACHE
>
> The last_level_cache_is_valid() use cacheinfo->{attributes, fw_token} to
> test it's valid or not, but populate_cache_leaves() will only reset
> LLC's type, so we won't try to re-setup LLC's type and leave it
> CACHE_TYPE_NOCACHE and won't export it through sysfs.
>
> This patch tries to fix this by not re-populating the cache leaves if
> the LLC is valid.
>
> Fixes: 5944ce092b97 ("arch_topology: Build cacheinfo from primary CPU")
> Signed-off-by: Yicong Yang <yangyicong@...ilicon.com>
> ---
> drivers/base/cacheinfo.c | 8 ++++++++
> 1 file changed, 8 insertions(+)
>
> diff --git a/drivers/base/cacheinfo.c b/drivers/base/cacheinfo.c
> index f6573c335f4c..d65f169d36dd 100644
> --- a/drivers/base/cacheinfo.c
> +++ b/drivers/base/cacheinfo.c
> @@ -473,6 +473,13 @@ int detect_cache_attributes(unsigned int cpu)
> return ret;
>
> populate_leaves:
> + /*
> + * If LLC is valid the cache leaves were already populated so just go to
> + * update the cpu map.
> + */
> + if (last_level_cache_is_valid(cpu))
> + goto update_cpu_map;
> +
> /*
> * populate_cache_leaves() may completely setup the cache leaves and
> * shared_cpu_map or it may leave it partially setup.
> @@ -481,6 +488,7 @@ int detect_cache_attributes(unsigned int cpu)
> if (ret)
> goto free_ci;
>
> +update_cpu_map:
Maybe just a suggestion about the code itself,
it should be possible to replace the 'goto' by an 'if' condition.
(Similarly, the 'populate_leaves:' label could have been avoided.)
> /*
> * For systems using DT for cache hierarchy, fw_token
> * and shared_cpu_map will be set up here only if they are
Thanks for the fix,
Pierre
Powered by blists - more mailing lists