| 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
| ||
|
Message-ID: <56ee1d92-28ee-81cb-9c41-6ca7ea6556b0@bytedance.com>
Date: Thu, 27 Jul 2023 16:55:27 +0800
From: Qi Zheng <zhengqi.arch@...edance.com>
To: Damien Le Moal <dlemoal@...nel.org>
Cc: linux-kernel@...r.kernel.org, linux-mm@...ck.org, x86@...nel.org,
kvm@...r.kernel.org, xen-devel@...ts.xenproject.org,
linux-erofs@...ts.ozlabs.org, linux-f2fs-devel@...ts.sourceforge.net,
cluster-devel@...hat.com, linux-nfs@...r.kernel.org,
linux-mtd@...ts.infradead.org, rcu@...r.kernel.org, netdev@...r.kernel.org,
dri-devel@...ts.freedesktop.org, linux-arm-msm@...r.kernel.org,
dm-devel@...hat.com, linux-raid@...r.kernel.org,
linux-bcache@...r.kernel.org, virtualization@...ts.linux-foundation.org,
linux-fsdevel@...r.kernel.org, linux-ext4@...r.kernel.org,
linux-xfs@...r.kernel.org, linux-btrfs@...r.kernel.org,
Muchun Song <songmuchun@...edance.com>, akpm@...ux-foundation.org,
david@...morbit.com, tkhai@...ru, vbabka@...e.cz, roman.gushchin@...ux.dev,
djwong@...nel.org, brauner@...nel.org, paulmck@...nel.org, tytso@....edu,
steven.price@....com, cel@...nel.org, senozhatsky@...omium.org,
yujie.liu@...el.com, gregkh@...uxfoundation.org, muchun.song@...ux.dev
Subject: Re: [PATCH v3 28/49] dm zoned: dynamically allocate the dm-zoned-meta
shrinker
Hi,
On 2023/7/27 16:30, Damien Le Moal wrote:
> On 7/27/23 17:04, Qi Zheng wrote:
>> In preparation for implementing lockless slab shrink, use new APIs to
>> dynamically allocate the dm-zoned-meta shrinker, so that it can be freed
>> asynchronously using kfree_rcu(). Then it doesn't need to wait for RCU
>> read-side critical section when releasing the struct dmz_metadata.
>>
>> Signed-off-by: Qi Zheng <zhengqi.arch@...edance.com>
>> Reviewed-by: Muchun Song <songmuchun@...edance.com>
>> ---
>> drivers/md/dm-zoned-metadata.c | 28 ++++++++++++++++------------
>> 1 file changed, 16 insertions(+), 12 deletions(-)
>>
>> diff --git a/drivers/md/dm-zoned-metadata.c b/drivers/md/dm-zoned-metadata.c
>> index 9d3cca8e3dc9..0bcb26a43578 100644
>> --- a/drivers/md/dm-zoned-metadata.c
>> +++ b/drivers/md/dm-zoned-metadata.c
>> @@ -187,7 +187,7 @@ struct dmz_metadata {
>> struct rb_root mblk_rbtree;
>> struct list_head mblk_lru_list;
>> struct list_head mblk_dirty_list;
>> - struct shrinker mblk_shrinker;
>> + struct shrinker *mblk_shrinker;
>>
>> /* Zone allocation management */
>> struct mutex map_lock;
>> @@ -615,7 +615,7 @@ static unsigned long dmz_shrink_mblock_cache(struct dmz_metadata *zmd,
>> static unsigned long dmz_mblock_shrinker_count(struct shrinker *shrink,
>> struct shrink_control *sc)
>> {
>> - struct dmz_metadata *zmd = container_of(shrink, struct dmz_metadata, mblk_shrinker);
>> + struct dmz_metadata *zmd = shrink->private_data;
>>
>> return atomic_read(&zmd->nr_mblks);
>> }
>> @@ -626,7 +626,7 @@ static unsigned long dmz_mblock_shrinker_count(struct shrinker *shrink,
>> static unsigned long dmz_mblock_shrinker_scan(struct shrinker *shrink,
>> struct shrink_control *sc)
>> {
>> - struct dmz_metadata *zmd = container_of(shrink, struct dmz_metadata, mblk_shrinker);
>> + struct dmz_metadata *zmd = shrink->private_data;
>> unsigned long count;
>>
>> spin_lock(&zmd->mblk_lock);
>> @@ -2936,19 +2936,23 @@ int dmz_ctr_metadata(struct dmz_dev *dev, int num_dev,
>> */
>> zmd->min_nr_mblks = 2 + zmd->nr_map_blocks + zmd->zone_nr_bitmap_blocks * 16;
>> zmd->max_nr_mblks = zmd->min_nr_mblks + 512;
>> - zmd->mblk_shrinker.count_objects = dmz_mblock_shrinker_count;
>> - zmd->mblk_shrinker.scan_objects = dmz_mblock_shrinker_scan;
>> - zmd->mblk_shrinker.seeks = DEFAULT_SEEKS;
>>
>> /* Metadata cache shrinker */
>> - ret = register_shrinker(&zmd->mblk_shrinker, "dm-zoned-meta:(%u:%u)",
>> - MAJOR(dev->bdev->bd_dev),
>> - MINOR(dev->bdev->bd_dev));
>> - if (ret) {
>> - dmz_zmd_err(zmd, "Register metadata cache shrinker failed");
>> + zmd->mblk_shrinker = shrinker_alloc(0, "dm-zoned-meta:(%u:%u)",
>> + MAJOR(dev->bdev->bd_dev),
>> + MINOR(dev->bdev->bd_dev));
>> + if (!zmd->mblk_shrinker) {
>> + dmz_zmd_err(zmd, "Allocate metadata cache shrinker failed");
>
> ret is not set here, so dmz_ctr_metadata() will return success. You need to add:
> ret = -ENOMEM;
> or something.
Indeed, will fix.
>> goto err;
>> }
>>
>> + zmd->mblk_shrinker->count_objects = dmz_mblock_shrinker_count;
>> + zmd->mblk_shrinker->scan_objects = dmz_mblock_shrinker_scan;
>> + zmd->mblk_shrinker->seeks = DEFAULT_SEEKS;
>> + zmd->mblk_shrinker->private_data = zmd;
>> +
>> + shrinker_register(zmd->mblk_shrinker);
>
> I fail to see how this new shrinker API is better... Why isn't there a
> shrinker_alloc_and_register() function ? That would avoid adding all this code
> all over the place as the new API call would be very similar to the current
> shrinker_register() call with static allocation.
In some registration scenarios, memory needs to be allocated in advance.
So we continue to use the previous prealloc/register_prepared()
algorithm. The shrinker_alloc_and_register() is just a helper function
that combines the two, and this increases the number of APIs that
shrinker exposes to the outside, so I choose not to add this helper.
Thanks,
Qi
>
>> +
>> dmz_zmd_info(zmd, "DM-Zoned metadata version %d", zmd->sb_version);
>> for (i = 0; i < zmd->nr_devs; i++)
>> dmz_print_dev(zmd, i);
>> @@ -2995,7 +2999,7 @@ int dmz_ctr_metadata(struct dmz_dev *dev, int num_dev,
>> */
>> void dmz_dtr_metadata(struct dmz_metadata *zmd)
>> {
>> - unregister_shrinker(&zmd->mblk_shrinker);
>> + shrinker_free(zmd->mblk_shrinker);
>> dmz_cleanup_metadata(zmd);
>> kfree(zmd);
>> }
>
Powered by blists - more mailing lists