Re: [PATCH v8 06/11] btrfs: implement RST version of scrub

[Qu Wenruo]

On 2023/9/11 22:22, Johannes Thumshirn wrote:
> A filesystem that uses the RAID stripe tree for logical to physical
> address translation can't use the regular scrub path, that reads all
> stripes and then checks if a sector is unused afterwards.
>
> When using the RAID stripe tree, this will result in lookup errors, as the
> stripe tree doesn't know the requested logical addresses.
>
> Instead, look up stripes that are backed by the extent bitmap.
>
> Signed-off-by: Johannes Thumshirn <johannes.thumshirn@xxxxxxx>
> ---
>   fs/btrfs/scrub.c | 56 ++++++++++++++++++++++++++++++++++++++++++++++++++++++++
>   1 file changed, 56 insertions(+)
>
> diff --git a/fs/btrfs/scrub.c b/fs/btrfs/scrub.c
> index f16220ce5fba..5101e0a3f83e 100644
> --- a/fs/btrfs/scrub.c
> +++ b/fs/btrfs/scrub.c
> @@ -23,6 +23,7 @@
>   #include "accessors.h"
>   #include "file-item.h"
>   #include "scrub.h"
> +#include "raid-stripe-tree.h"
>
>   /*
>    * This is only the first step towards a full-features scrub. It reads all
> @@ -1634,6 +1635,56 @@ static void scrub_reset_stripe(struct scrub_stripe *stripe)
>   	}
>   }
>
> +static void scrub_submit_extent_sector_read(struct scrub_ctx *sctx,
> +					    struct scrub_stripe *stripe)
> +{
> +	struct btrfs_fs_info *fs_info = stripe->bg->fs_info;
> +	struct btrfs_bio *bbio = NULL;
> +	int mirror = stripe->mirror_num;
> +	int i;
> +
> +	atomic_inc(&stripe->pending_io);
> +
> +	for_each_set_bit(i, &stripe->extent_sector_bitmap, stripe->nr_sectors) {
> +		struct page *page;
> +		int pgoff;
> +
> +		page = scrub_stripe_get_page(stripe, i);
> +		pgoff = scrub_stripe_get_page_offset(stripe, i);
> +
> +		/* The current sector cannot be merged, submit the bio. */
> +		if (bbio &&
> +		    ((i > 0 && !test_bit(i - 1, &stripe->extent_sector_bitmap)) ||
> +		     bbio->bio.bi_iter.bi_size >= BTRFS_STRIPE_LEN)) {
> +			ASSERT(bbio->bio.bi_iter.bi_size);
> +			atomic_inc(&stripe->pending_io);
> +			btrfs_submit_bio(bbio, mirror);
> +			bbio = NULL;
> +		}
> +
> +		if (!bbio) {
> +			bbio = btrfs_bio_alloc(stripe->nr_sectors, REQ_OP_READ,
> +				fs_info, scrub_read_endio, stripe);
> +			bbio->bio.bi_iter.bi_sector = (stripe->logical +
> +				(i << fs_info->sectorsize_bits)) >> SECTOR_SHIFT;
> +		}
> +
> +		__bio_add_page(&bbio->bio, page, fs_info->sectorsize, pgoff);
> +	}
> +
> +	if (bbio) {
> +		ASSERT(bbio->bio.bi_iter.bi_size);
> +		atomic_inc(&stripe->pending_io);
> +		btrfs_submit_bio(bbio, mirror);

Since RST is looked up during btrfs_submit_bio() (to be more accurate,
set_io_stripe()), and I just checked there is no special requirement to
make btrfs to lookup using commit root.

This means we can have a problem that extent items and RST are out-of-sync.

For scrub, all the extent items are searched using commit root, but
btrfs_get_raid_extent_offset() is only using current root.
Thus you would got some problems during fsstress and scrub.


We need some way to distinguish scrub bbio from regular ones (which is a
completely new requirement).
For now only scrub doesn't initialize bbio->inode, thus it can be used
to do the distinguish (at least for now).

Thanks,
Qu
> +	}
> +
> +	if (atomic_dec_and_test(&stripe->pending_io)) {
> +		wake_up(&stripe->io_wait);
> +		INIT_WORK(&stripe->work, scrub_stripe_read_repair_worker);
> +		queue_work(stripe->bg->fs_info->scrub_workers, &stripe->work);
> +	}
> +}
> +
>   static void scrub_submit_initial_read(struct scrub_ctx *sctx,
>   				      struct scrub_stripe *stripe)
>   {
> @@ -1645,6 +1696,11 @@ static void scrub_submit_initial_read(struct scrub_ctx *sctx,
>   	ASSERT(stripe->mirror_num > 0);
>   	ASSERT(test_bit(SCRUB_STRIPE_FLAG_INITIALIZED, &stripe->state));
>
> +	if (btrfs_need_stripe_tree_update(fs_info, stripe->bg->flags)) {
> +		scrub_submit_extent_sector_read(sctx, stripe);
> +		return;
> +	}
> +
>   	bbio = btrfs_bio_alloc(SCRUB_STRIPE_PAGES, REQ_OP_READ, fs_info,
>   			       scrub_read_endio, stripe);