platform/kernel/linux-rpi.git
15 months agobtrfs: avoid repetitive define BTRFS_FEATURE_INCOMPAT_SUPP
Anand Jain [Mon, 6 Mar 2023 13:28:09 +0000 (21:28 +0800)]
btrfs: avoid repetitive define BTRFS_FEATURE_INCOMPAT_SUPP

BTRFS_FEATURE_INCOMPAT_SUPP is defined twice, once under
CONFIG_BTRFS_DEBUG and once without it, resulting in repetitive code. The
reason for this is to add experimental features under CONFIG_BTRFS_DEBUG.

To avoid repetitive code, add a common list BTRFS_FEATURE_INCOMPAT_SUPP_STABLE,
and append experimental features only under CONFIG_BTRFS_DEBUG.

Signed-off-by: Anand Jain <anand.jain@oracle.com>
Reviewed-by: David Sterba <dsterba@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
15 months agobtrfs: scrub: remove root and csum_root arguments from scrub_simple_mirror()
Qu Wenruo [Mon, 16 Jan 2023 07:04:13 +0000 (15:04 +0800)]
btrfs: scrub: remove root and csum_root arguments from scrub_simple_mirror()

We don't need to pass the roots as arguments, reading them from the
rb-tree is cheap.  Thus there is really not much need to pre-fetch it
and pass it all the way from scrub_stripe().

And we already have more than enough arguments in scrub_simple_mirror()
and scrub_simple_stripe(), it's better to remove them and only grab
those roots in scrub_simple_mirror().

Signed-off-by: Qu Wenruo <wqu@suse.com>
Reviewed-by: David Sterba <dsterba@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
15 months agobtrfs: scrub: remove unused path inside scrub_stripe()
Qu Wenruo [Mon, 16 Jan 2023 07:04:12 +0000 (15:04 +0800)]
btrfs: scrub: remove unused path inside scrub_stripe()

The variable @path is no longer passed into any call sites after commit
18d30ab96149 ("btrfs: scrub: use scrub_simple_mirror() to handle RAID56
data stripe scrub"), thus we can remove the variable completely.

Signed-off-by: Qu Wenruo <wqu@suse.com>
Reviewed-by: David Sterba <dsterba@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
15 months agobtrfs: do not use replace target device as an extra mirror
Qu Wenruo [Wed, 22 Feb 2023 07:47:40 +0000 (15:47 +0800)]
btrfs: do not use replace target device as an extra mirror

[BUG]
Currently btrfs can use dev-replace device as an extra mirror for
read-repair.  But it can lead to NODATASUM corruption in the following
case:

 There is a RAID1 data chunk, and dev-replace is running from
 dev2 to dev0.

 |//| = Replaced data
          X       X+1MB     X+2MB
  Dev 2:  |       |         |           <- Source dev
  Dev 0:  |///////|         |           <- Target dev

Then a read on dev 2 X+2MB happens.
And something wrong happened inside devid 2, causing an -EIO.

In that case, read-repair would try the next mirror, and since we can
use target device as an extra mirror, we will use that mirror instead.

But unfortunately since the read is beyond the current replace cursor,
we should not trust it at all, what we get would be just uninitialized
garbage.

But if this read is for NODATASUM range, then we just trust them and
cause data corruption.

[CAUSE]
We used to have some checks to make sure we only return such extra
mirror when the range is before our left cursor.

The first commit introducing this behavior is ad6d620e2a57 ("Btrfs:
allow repair code to include target disk when searching mirrors").

But later a fix, 22ab04e814f4 ("Btrfs: fix race between device replace
and chunk allocation") changed the behavior, to always let
btrfs_map_block() include the extra mirror to address a race in
dev-replace which can cause missing writes to target device.

This means, we lose the tracking of cursor for the extra mirror, thus
can lead to above corruption.

[FIX]
The extra mirror is never a reliable one, at the beginning of
dev-replace, the reliability is zero, while only at the end of the
replace it's a fully reliable mirror.

We either do the complex tracking, or never trust it.

IMHO it's much easier to maintain if we don't trust it at all, and the
extra mirror can only benefit for a limited period of time (during
replace).

Thus this patch would completely remove the ability to use target device
as an extra mirror.

Signed-off-by: Qu Wenruo <wqu@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
15 months agobtrfs: open_ctree() error handling cleanup
Qu Wenruo [Tue, 28 Feb 2023 00:44:30 +0000 (08:44 +0800)]
btrfs: open_ctree() error handling cleanup

Currently open_ctree() still uses two variables for error handling, err
and ret. This can be confusing and missing some errors and does not
conform to current coding style.

This patch will fix the problems by:

- Use only ret for error handling

- Add proper ret assignment
  Originally we rely on the default value (-EINVAL) of err to handle
  errors, but that doesn't really reflects the error.
  This will change it use the correct error number for the following
  call sites:

  * subpage_info allocation
  * btrfs_free_extra_devids()
  * btrfs_check_rw_degradable()
  * cleaner_kthread allocation
  * transaction_kthread allocation

- Add an extra ASSERT()
  To make sure we error out instead of returning 0.

Reviewed-by: Anand Jain <anand.jain@oracle.com>
Signed-off-by: Qu Wenruo <wqu@suse.com>
Reviewed-by: David Sterba <dsterba@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
15 months agobtrfs: cleanup the main loop in btrfs_lookup_bio_sums
Christoph Hellwig [Wed, 22 Feb 2023 17:07:02 +0000 (09:07 -0800)]
btrfs: cleanup the main loop in btrfs_lookup_bio_sums

Introduce a bio_offset variable for the current offset into the bio
instead of recalculating it over and over.   Remove the now only used
once search_len and sector_offset variables, and reduce the scope for
count and cur_disk_bytenr.

Reviewed-by: Johannes Thumshirn <johannes.thumshirn@wdc.com>
Signed-off-by: Christoph Hellwig <hch@lst.de>
Signed-off-by: David Sterba <dsterba@suse.com>
15 months agobtrfs: remove search_file_offset_in_bio
Christoph Hellwig [Wed, 22 Feb 2023 17:07:01 +0000 (09:07 -0800)]
btrfs: remove search_file_offset_in_bio

There is no need to search for a file offset in a bio, it is now always
provided in bbio->file_offset (set at bio allocation time since
0d495430db8d ("btrfs: set bbio->file_offset in alloc_new_bio")).  Just
use that with the offset into the bio.

Reviewed-by: Anand Jain <anand.jain@oracle.com>
Reviewed-by: Johannes Thumshirn <johannes.thumshirn@wdc.com>
Signed-off-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: David Sterba <dsterba@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
15 months agobtrfs: sink calc_bio_boundaries into its only caller
Johannes Thumshirn [Tue, 21 Feb 2023 16:21:04 +0000 (08:21 -0800)]
btrfs: sink calc_bio_boundaries into its only caller

Nowadays calc_bio_boundaries() is a relatively simple function that only
guarantees the one bio equals to one ordered extent rule for uncompressed
Zone Append bios.

Sink it into it's only caller alloc_new_bio().

Reviewed-by: Christoph Hellwig <hch@lst.de>
Signed-off-by: Johannes Thumshirn <johannes.thumshirn@wdc.com>
Reviewed-by: David Sterba <dsterba@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
15 months agobtrfs: simplify main loop in submit_extent_page
Christoph Hellwig [Mon, 27 Feb 2023 15:17:04 +0000 (08:17 -0700)]
btrfs: simplify main loop in submit_extent_page

bio_add_page always adds either the entire range passed to it or nothing.
Based on that btrfs_bio_add_page can only return a length smaller than
the passed in one when hitting the ordered extent limit, which can only
happen for writes.  Given that compressed writes never even use this code
path, this means that all the special cases for compressed extent offset
handling are dead code.

Reflow submit_extent_page to take advantage of this by inlining
btrfs_bio_add_page and handling the ordered extent limit by decrementing
it for each added range and thus significantly simplifying the loop.

Reviewed-by: Johannes Thumshirn <johannes.thumshirn@wdc.com>
Signed-off-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: David Sterba <dsterba@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
15 months agobtrfs: check for contiguity in submit_extent_page
Christoph Hellwig [Mon, 27 Feb 2023 15:17:03 +0000 (08:17 -0700)]
btrfs: check for contiguity in submit_extent_page

Different loop iterations in btrfs_bio_add_page not only have the same
contiguity parameters, but also any non-initial operation operates on a
fresh bio anyway.

Factor out the contiguity check into a new btrfs_bio_is_contig and only
call it once in submit_extent_page before descending into the
bio_add_page loop.

Reviewed-by: Johannes Thumshirn <johannes.thumshirn@wdc.com>
Signed-off-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: David Sterba <dsterba@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
15 months agobtrfs: simplify the error handling in __extent_writepage_io
Christoph Hellwig [Mon, 27 Feb 2023 15:17:02 +0000 (08:17 -0700)]
btrfs: simplify the error handling in __extent_writepage_io

Remove the has_error and saved_ret variables, and just jump to a goto
label for error handling from the only place returning an error from the
main loop.

Reviewed-by: Johannes Thumshirn <johannes.thumshirn@wdc.com>
Signed-off-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: David Sterba <dsterba@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
15 months agobtrfs: remove the submit_extent_page return value
Christoph Hellwig [Mon, 27 Feb 2023 15:17:01 +0000 (08:17 -0700)]
btrfs: remove the submit_extent_page return value

submit_extent_page always returns 0 since commit d5e4377d5051 ("btrfs:
split zone append bios in btrfs_submit_bio").  Change it to a void return
type and remove all the unreachable error handling code in the callers.

Reviewed-by: Johannes Thumshirn <johannes.thumshirn@wdc.com>
Signed-off-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: David Sterba <dsterba@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
15 months agobtrfs: remove the compress_type argument to submit_extent_page
Christoph Hellwig [Mon, 27 Feb 2023 15:17:00 +0000 (08:17 -0700)]
btrfs: remove the compress_type argument to submit_extent_page

Update the compress_type in the btrfs_bio_ctrl after forcing out the
previous bio in btrfs_do_readpage, so that alloc_new_bio can just use
the compress_type member in struct btrfs_bio_ctrl instead of passing the
same information redundantly as a function argument.

Reviewed-by: Johannes Thumshirn <johannes.thumshirn@wdc.com>
Signed-off-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: David Sterba <dsterba@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
15 months agobtrfs: rename the this_bio_flag variable in btrfs_do_readpage
Christoph Hellwig [Mon, 27 Feb 2023 15:16:59 +0000 (08:16 -0700)]
btrfs: rename the this_bio_flag variable in btrfs_do_readpage

Rename this_bio_flag to compress_type to match the surrounding code
and better document the intent.  Also use the proper enum type instead
of unsigned long.

Reviewed-by: Johannes Thumshirn <johannes.thumshirn@wdc.com>
Reviewed-by: Qu Wenruo <wqu@suse.com>
Signed-off-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: David Sterba <dsterba@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
15 months agobtrfs: move the compress_type check out of btrfs_bio_add_page
Christoph Hellwig [Mon, 27 Feb 2023 15:16:58 +0000 (08:16 -0700)]
btrfs: move the compress_type check out of btrfs_bio_add_page

The compress_type can only change on a per-extent basis.  So instead of
checking it for every page in btrfs_bio_add_page, do the check once in
btrfs_do_readpage, which is the only caller of btrfs_bio_add_page and
submit_extent_page that deals with compressed extents.

Signed-off-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: David Sterba <dsterba@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
15 months agobtrfs: add a wbc pointer to struct btrfs_bio_ctrl
Christoph Hellwig [Mon, 27 Feb 2023 15:16:57 +0000 (08:16 -0700)]
btrfs: add a wbc pointer to struct btrfs_bio_ctrl

Instead of passing down the wbc pointer the deep call chain, just
add it to the btrfs_bio_ctrl structure.

Reviewed-by: Johannes Thumshirn <johannes.thumshirn@wdc.com>
Signed-off-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: David Sterba <dsterba@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
15 months agobtrfs: remove the sync_io flag in struct btrfs_bio_ctrl
Christoph Hellwig [Mon, 27 Feb 2023 15:16:56 +0000 (08:16 -0700)]
btrfs: remove the sync_io flag in struct btrfs_bio_ctrl

The sync_io flag is equivalent to wbc->sync_mode == WB_SYNC_ALL, so
just check for that and remove the separate flag.

Reviewed-by: Johannes Thumshirn <johannes.thumshirn@wdc.com>
Signed-off-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: David Sterba <dsterba@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
15 months agobtrfs: store the bio opf in struct btrfs_bio_ctrl
Christoph Hellwig [Mon, 27 Feb 2023 15:16:55 +0000 (08:16 -0700)]
btrfs: store the bio opf in struct btrfs_bio_ctrl

The bio op and flags never change over the life time of a bio_ctrl,
so move it in there instead of passing it down the deep call chain
all the way down to alloc_new_bio.

Reviewed-by: Johannes Thumshirn <johannes.thumshirn@wdc.com>
Reviewed-by: Qu Wenruo <wqu@suse.com>
Signed-off-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: David Sterba <dsterba@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
15 months agobtrfs: remove the force_bio_submit to submit_extent_page
Christoph Hellwig [Mon, 27 Feb 2023 15:16:54 +0000 (08:16 -0700)]
btrfs: remove the force_bio_submit to submit_extent_page

If force_bio_submit, submit_extent_page simply calls submit_one_bio as
the first thing.  This can just be moved to the only caller that sets
force_bio_submit to true.

Reviewed-by: Johannes Thumshirn <johannes.thumshirn@wdc.com>
Signed-off-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: David Sterba <dsterba@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
15 months agobtrfs: don't set force_bio_submit in read_extent_buffer_subpage
Christoph Hellwig [Mon, 27 Feb 2023 15:16:53 +0000 (08:16 -0700)]
btrfs: don't set force_bio_submit in read_extent_buffer_subpage

When read_extent_buffer_subpage calls submit_extent_page, it does
so on a freshly initialized btrfs_bio_ctrl structure that can't have
a valid bio to submit.  Clear the force_bio_submit parameter to false
as there is nothing to submit.

Reviewed-by: Johannes Thumshirn <johannes.thumshirn@wdc.com>
Signed-off-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: David Sterba <dsterba@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
15 months agobtrfs: open code btrfs_bin_search()
Anand Jain [Fri, 24 Feb 2023 03:31:26 +0000 (11:31 +0800)]
btrfs: open code btrfs_bin_search()

btrfs_bin_search() is a simple wrapper that searches for the whole slots
by calling btrfs_generic_bin_search() with the starting slot/first_slot
preset to 0.

This simple wrapper can be open coded as btrfs_bin_search().

Signed-off-by: Anand Jain <anand.jain@oracle.com>
Reviewed-by: David Sterba <dsterba@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
15 months agobtrfs: dev-replace: properly follow its read mode
Qu Wenruo [Wed, 22 Feb 2023 07:04:37 +0000 (15:04 +0800)]
btrfs: dev-replace: properly follow its read mode

[BUG]
Although dev replace ioctl has a way to specify the mode on whether we
should read from the source device, it's not properly followed.

 # mkfs.btrfs -f -d raid1 -m raid1 $dev1 $dev2
 # mount $dev1 $mnt
 # xfs_io -f -c "pwrite 0 32M" $mnt/file
 # sync
 # btrfs replace start -r -f 1 $dev3 $mnt

And one extra trace is added to scrub_submit(), showing the detail about
the bio:

  btrfs-11569 [005] ...  37.0270: scrub_submit.part.0: devid=1 logical=22036480 phy=22036480 len=16384
  btrfs-11569 [005] ...  37.0273: scrub_submit.part.0: devid=1 logical=30457856 phy=30457856 len=32768
  btrfs-11569 [005] ...  37.0274: scrub_submit.part.0: devid=1 logical=30507008 phy=30507008 len=49152
  btrfs-11569 [005] ...  37.0274: scrub_submit.part.0: devid=1 logical=30605312 phy=30605312 len=32768
  btrfs-11569 [005] ...  37.0275: scrub_submit.part.0: devid=1 logical=30703616 phy=30703616 len=65536
  btrfs-11569 [005] ...  37.0281: scrub_submit.part.0: devid=1 logical=298844160 phy=298844160 len=131072
  ...
  btrfs-11569 [005] ...  37.0762: scrub_submit.part.0: devid=1 logical=322961408 phy=322961408 len=131072
  btrfs-11569 [005] ...  37.0762: scrub_submit.part.0: devid=1 logical=323092480 phy=323092480 len=131072

One can see that all the reads are submitted to devid 1, even if we have
specified "-r" option to avoid reading from the source device.

[CAUSE]
The dev-replace read mode is only set but not followed by scrub code at
all.  In fact, only common read path is properly following the read
mode, but scrub itself has its own read path, thus not following the
mode.

[FIX]
Here we enhance scrub_find_good_copy() to also follow the read mode.

The idea is pretty simple, in the first loop, we avoid the following
devices:

- Missing devices
  This is the existing condition

- The source device if the replace wants to avoid it.

And if above loop found no candidate (e.g. replace a single device),
then we discard the 2nd condition, and try again.

Since we're here, also enhance the function scrub_find_good_copy() by:

- Remove the forward declaration

- Makes it return int
  To indicates errors, e.g. no good mirror found.

- Add extra error messages

Now with the same trace, "btrfs replace start -r" works as expected:

  btrfs-1213 [000] ...  991.9059: scrub_submit.part.0: devid=2 logical=22036480 phy=1064960 len=16384
  btrfs-1213 [000] ...  991.9062: scrub_submit.part.0: devid=2 logical=30457856 phy=9486336 len=32768
  btrfs-1213 [000] ...  991.9063: scrub_submit.part.0: devid=2 logical=30507008 phy=9535488 len=49152
  btrfs-1213 [000] ...  991.9064: scrub_submit.part.0: devid=2 logical=30605312 phy=9633792 len=32768
  btrfs-1213 [000] ...  991.9065: scrub_submit.part.0: devid=2 logical=30703616 phy=9732096 len=65536
  btrfs-1213 [000] ...  991.9073: scrub_submit.part.0: devid=2 logical=298844160 phy=277872640 len=131072
  btrfs-1213 [000] ...  991.9075: scrub_submit.part.0: devid=2 logical=298975232 phy=278003712 len=131072
  btrfs-1213 [000] ...  991.9078: scrub_submit.part.0: devid=2 logical=299106304 phy=278134784 len=131072
  ...
  btrfs-1213 [000] ...  991.9474: scrub_submit.part.0: devid=2 logical=318504960 phy=297533440 len=131072
  btrfs-1213 [000] ...  991.9476: scrub_submit.part.0: devid=2 logical=318636032 phy=297664512 len=131072
  btrfs-1213 [000] ...  991.9479: scrub_submit.part.0: devid=2 logical=318767104 phy=297795584 len=131072

Reviewed-by: Anand Jain <anand.jain@oracle.com>
Signed-off-by: Qu Wenruo <wqu@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
15 months agobtrfs: fold finish_compressed_bio_write into btrfs_finish_compressed_write_work
Christoph Hellwig [Fri, 10 Feb 2023 07:48:41 +0000 (08:48 +0100)]
btrfs: fold finish_compressed_bio_write into btrfs_finish_compressed_write_work

Fold finish_compressed_bio_write into its only caller as there is no
reason to keep them separate.

Reviewed-by: Johannes Thumshirn <johannes.thumshirn@wdc.com>
Signed-off-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: David Sterba <dsterba@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
15 months agobtrfs: don't clear page->mapping in btrfs_free_compressed_pages
Christoph Hellwig [Fri, 10 Feb 2023 07:48:40 +0000 (08:48 +0100)]
btrfs: don't clear page->mapping in btrfs_free_compressed_pages

No one ever set ->mapping on these pages, so don't bother clearing it.

Reviewed-by: Johannes Thumshirn <johannes.thumshirn@wdc.com>
Signed-off-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: David Sterba <dsterba@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
15 months agobtrfs: factor out a btrfs_free_compressed_pages helper
Christoph Hellwig [Fri, 10 Feb 2023 07:48:39 +0000 (08:48 +0100)]
btrfs: factor out a btrfs_free_compressed_pages helper

Share the code to free the compressed pages and the array to hold them
into a common helper.

Reviewed-by: Johannes Thumshirn <johannes.thumshirn@wdc.com>
Signed-off-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: David Sterba <dsterba@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
15 months agobtrfs: factor out a btrfs_add_compressed_bio_pages helper
Christoph Hellwig [Fri, 10 Feb 2023 07:48:38 +0000 (08:48 +0100)]
btrfs: factor out a btrfs_add_compressed_bio_pages helper

Factor out a common helper to add the compressed_bio pages to the
bio that is shared by the compressed read and write path.

Reviewed-by: Johannes Thumshirn <johannes.thumshirn@wdc.com>
Signed-off-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: David Sterba <dsterba@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
15 months agobtrfs: use the bbio file offset in add_ra_bio_pages
Christoph Hellwig [Fri, 10 Feb 2023 07:48:37 +0000 (08:48 +0100)]
btrfs: use the bbio file offset in add_ra_bio_pages

struct btrfs_bio now has a file_offset field set up by all submitters.
Use that value combined with the bio size in add_ra_bio_pages to
calculate the last offset in the bio.

Reviewed-by: Johannes Thumshirn <johannes.thumshirn@wdc.com>
Signed-off-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: David Sterba <dsterba@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
15 months agobtrfs: use the bbio file offset in btrfs_submit_compressed_read
Christoph Hellwig [Fri, 10 Feb 2023 07:48:36 +0000 (08:48 +0100)]
btrfs: use the bbio file offset in btrfs_submit_compressed_read

struct btrfs_bio now has a file_offset field set up by all submitters.
Use that in btrfs_submit_compressed_read instead of recalculating the
value.

Reviewed-by: Johannes Thumshirn <johannes.thumshirn@wdc.com>
Signed-off-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: David Sterba <dsterba@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
15 months agobtrfs: remove redundant free_extent_map in btrfs_submit_compressed_read
Christoph Hellwig [Fri, 10 Feb 2023 07:48:35 +0000 (08:48 +0100)]
btrfs: remove redundant free_extent_map in btrfs_submit_compressed_read

em can't be non-NULL after the free_extent_map label.  Also remove
the now pointless clearing of em to NULL after freeing it.

Reviewed-by: Johannes Thumshirn <johannes.thumshirn@wdc.com>
Signed-off-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: David Sterba <dsterba@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
15 months agobtrfs: embed a btrfs_bio into struct compressed_bio
Christoph Hellwig [Fri, 10 Feb 2023 07:48:34 +0000 (08:48 +0100)]
btrfs: embed a btrfs_bio into struct compressed_bio

Embed a btrfs_bio into struct compressed_bio.  This avoids potential
(so far theoretical) deadlocks due to nesting of btrfs_bioset allocations
for the original read bio and the compressed bio, and avoids an extra
memory allocation in the I/O path.

Reviewed-by: Johannes Thumshirn <johannes.thumshirn@wdc.com>
Signed-off-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: David Sterba <dsterba@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
15 months agobtrfs: replace btrfs_io_context::raid_map with a fixed u64 value
Qu Wenruo [Fri, 17 Feb 2023 05:37:03 +0000 (13:37 +0800)]
btrfs: replace btrfs_io_context::raid_map with a fixed u64 value

In btrfs_io_context structure, we have a pointer raid_map, which
indicates the logical bytenr for each stripe.

But considering we always call sort_parity_stripes(), the result
raid_map[] is always sorted, thus raid_map[0] is always the logical
bytenr of the full stripe.

So why we waste the space and time (for sorting) for raid_map?

This patch will replace btrfs_io_context::raid_map with a single u64
number, full_stripe_start, by:

- Replace btrfs_io_context::raid_map with full_stripe_start

- Replace call sites using raid_map[0] to use full_stripe_start

- Replace call sites using raid_map[i] to compare with nr_data_stripes.

The benefits are:

- Less memory wasted on raid_map
  It's sizeof(u64) * num_stripes vs sizeof(u64).
  It'll always save at least one u64, and the benefit grows larger with
  num_stripes.

- No more weird alloc_btrfs_io_context() behavior
  As there is only one fixed size + one variable length array.

Signed-off-by: Qu Wenruo <wqu@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
15 months agobtrfs: use an efficient way to represent source of duplicated stripes
Qu Wenruo [Tue, 7 Feb 2023 04:26:14 +0000 (12:26 +0800)]
btrfs: use an efficient way to represent source of duplicated stripes

For btrfs dev-replace, we have to duplicate writes to the source
device into the target device.

For non-RAID56, all writes into the same mapped ranges are sharing the
same content, thus they don't really need to bother anything.
(E.g. in btrfs_submit_bio() for non-RAID56 range we just submit the
same write to all involved devices).

But for RAID56, all stripes contain different content, thus we must
have a clear mapping of which stripe is duplicated from which original
stripe.

Currently we use a complex way using tgtdev_map[] array, e.g:

 num_tgtdevs = 1
 tgtdev_map[0] = 0    <- Means stripes[0] is not involved in replace.
 tgtdev_map[1] = 3    <- Means stripes[1] is involved in replace,
 and it's duplicated to stripes[3].
 tgtdev_map[2] = 0    <- Means stripes[2] is not involved in replace.

But this is wasting some space, and ignores one important thing for
dev-replace, there is at most one running replace.

Thus we can change it to a fixed array to represent the mapping:

 replace_nr_stripes = 1
 replace_stripe_src = 1    <- Means stripes[1] is involved in replace.
      thus the extra stripe is a copy of
      stripes[1]

By this we can save some space for bioc on RAID56 chunks with many
devices.  And we get rid of one variable sized array from bioc.

Thus the patch involves the following changes:

- Replace @num_tgtdevs and @tgtdev_map[] with @replace_nr_stripes
  and @replace_stripe_src.

  @num_tgtdevs is just renamed to @replace_nr_stripes.
  While the mapping is completely changed.

- Add extra ASSERT()s for RAID56 code

- Only add two more extra stripes for dev-replace cases.
  As we have an upper limit on how many dev-replace stripes we can have.

- Unify the behavior of handle_ops_on_dev_replace()
  Previously handle_ops_on_dev_replace() go two different paths for
  WRITE and GET_READ_MIRRORS.
  Now unify them by always going the WRITE path first (with at most 2
  replace stripes), then if we're doing GET_READ_MIRRORS and we have 2
  extra stripes, just drop one stripe.

- Remove the @real_stripes argument from alloc_btrfs_io_context()
  As we don't need the old variable length array any more.

Signed-off-by: Qu Wenruo <wqu@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
15 months agobtrfs: reduce type width of btrfs_io_contexts
Qu Wenruo [Tue, 7 Feb 2023 04:26:13 +0000 (12:26 +0800)]
btrfs: reduce type width of btrfs_io_contexts

That structure is our ultimate object for all __btrfs_map_block()
related functions.  We have some hard to understand members, like
tgtdev_map, but without any comments.

This patch will improve the situation:

- Add extra comments for num_stripes, mirror_num, num_tgtdevs and
  tgtdev_map[]
  Especially for the last two members, add a dedicated (thus very long)
  comments for them, with example to explain it.

- Shrink those int members to u16.
  In fact our on-disk format is only using u16 for num_stripes, thus
  no need to use int at all.

Signed-off-by: Qu Wenruo <wqu@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
15 months agobtrfs: simplify the bioc argument for handle_ops_on_dev_replace()
Qu Wenruo [Tue, 7 Feb 2023 04:26:12 +0000 (12:26 +0800)]
btrfs: simplify the bioc argument for handle_ops_on_dev_replace()

There is no memory re-allocation for handle_ops_on_dev_replace(), thus
we don't need to pass a btrfs_io_context pointer.

Reviewed-by: Johannes Thumshirn <johannes.thumshirn@wdc.com>
Reviewed-by: Anand Jain <anand.jain@oracle.com>
Signed-off-by: Qu Wenruo <wqu@suse.com>
Reviewed-by: David Sterba <dsterba@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
15 months agobtrfs: reduce div64 calls by limiting the number of stripes of a chunk to u32
Qu Wenruo [Fri, 17 Feb 2023 05:36:59 +0000 (13:36 +0800)]
btrfs: reduce div64 calls by limiting the number of stripes of a chunk to u32

There are quite some div64 calls inside btrfs_map_block() and its
variants.

Such calls are for @stripe_nr, where @stripe_nr is the number of
stripes before our logical bytenr inside a chunk.

However we can eliminate such div64 calls by just reducing the width of
@stripe_nr from 64 to 32.

This can be done because our chunk size limit is already 10G, with fixed
stripe length 64K.
Thus a U32 is definitely enough to contain the number of stripes.

With such width reduction, we can get rid of slower div64, and extra
warning for certain 32bit arch.

This patch would do:

- Add a new tree-checker chunk validation on chunk length
  Make sure no chunk can reach 256G, which can also act as a bitflip
  checker.

- Reduce the width from u64 to u32 for @stripe_nr variables

- Replace unnecessary div64 calls with regular modulo and division
  32bit division and modulo are much faster than 64bit operations, and
  we are finally free of the div64 fear at least in those involved
  functions.

Signed-off-by: Qu Wenruo <wqu@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
15 months agobtrfs: replace map_lookup->stripe_len by BTRFS_STRIPE_LEN
Qu Wenruo [Fri, 17 Feb 2023 05:36:58 +0000 (13:36 +0800)]
btrfs: replace map_lookup->stripe_len by BTRFS_STRIPE_LEN

Currently btrfs doesn't support stripe lengths other than 64KiB.
This is already set in the tree-checker.

There is really no meaning to record that fixed value in map_lookup for
now, and can all be replaced with BTRFS_STRIPE_LEN.

Furthermore we can use the fix stripe length to do the following
optimization:

- Use BTRFS_STRIPE_LEN_SHIFT to replace some 64bit division
  Now we only need to do a right shift.

  And the value of BTRFS_STRIPE_LEN itself is already too large for bit
  shift, thus if we accidentally use BTRFS_STRIPE_LEN to do bit shift,
  a compiler warning would be triggered.

  Thus this bit shift optimization would be safe.

- Use BTRFS_STRIPE_LEN_MASK to calculate the offset inside a stripe

Reviewed-by: Johannes Thumshirn <johannes.thumshirn@wdc.com>
Reviewed-by: Christoph Hellwig <hch@lst.de>
Signed-off-by: Qu Wenruo <wqu@suse.com>
Reviewed-by: David Sterba <dsterba@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
15 months agobtrfs: move all btree inode initialization into btrfs_init_btree_inode
Christoph Hellwig [Sun, 19 Feb 2023 18:10:22 +0000 (19:10 +0100)]
btrfs: move all btree inode initialization into btrfs_init_btree_inode

Move the remaining code that deals with initializing the btree
inode into btrfs_init_btree_inode instead of splitting it between
that helpers and its only caller.

Reviewed-by: Johannes Thumshirn <johannes.thumshirn@wdc.com>
Reviewed-by: Anand Jain <anand.jain@oracle.com>
Signed-off-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: David Sterba <dsterba@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
15 months agobtrfs: switch search_file_offset_in_bio to return bool
Anand Jain [Fri, 10 Feb 2023 16:15:55 +0000 (00:15 +0800)]
btrfs: switch search_file_offset_in_bio to return bool

Function search_file_offset_in_bio() finds the file offset in the
file_offset_ret, and we use the return value to indicate if it is
successful, so use bool.

Reviewed-by: Christoph Hellwig <hch@lst.de>
Signed-off-by: Anand Jain <anand.jain@oracle.com>
Reviewed-by: David Sterba <dsterba@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
15 months agobtrfs: avoid reusing return variable in nested block in btrfs_lookup_bio_sums
Anand Jain [Fri, 10 Feb 2023 16:15:54 +0000 (00:15 +0800)]
btrfs: avoid reusing return variable in nested block in btrfs_lookup_bio_sums

The function btrfs_lookup_bio_sums() and a nested if statement declare
ret respectively as blk_status_t and int.

There is no need to store the return value of
search_file_offset_in_bio() to ret as this is a one-time call.

Reviewed-by: Christoph Hellwig <hch@lst.de>
Signed-off-by: Anand Jain <anand.jain@oracle.com>
Reviewed-by: David Sterba <dsterba@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
15 months agobtrfs: open code btrfs_csum_ptr
Johannes Thumshirn [Fri, 10 Feb 2023 10:50:08 +0000 (02:50 -0800)]
btrfs: open code btrfs_csum_ptr

Remove btrfs_csum_ptr() and fold it into it's only caller.

Reviewed-by: Anand Jain <anand.jain@oracle.com>
Reviewed-by: Christoph Hellwig <hch@lst.de>
Signed-off-by: Johannes Thumshirn <johannes.thumshirn@wdc.com>
Reviewed-by: David Sterba <dsterba@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
15 months agobtrfs: raid56: no need for irqsafe locking
Christoph Hellwig [Fri, 20 Jan 2023 07:46:57 +0000 (08:46 +0100)]
btrfs: raid56: no need for irqsafe locking

These days all the operations that take locks in the raid56.c code are
run from user context (mostly workqueues).  Drop all the irqsafe locking
that is not required any more.

Reviewed-by: Qu Wenruo <wqu@suse.com>
Signed-off-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: David Sterba <dsterba@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
15 months agobtrfs: abort the transaction if we get an error during snapshot drop
Josef Bacik [Tue, 7 Feb 2023 16:57:25 +0000 (11:57 -0500)]
btrfs: abort the transaction if we get an error during snapshot drop

We were seeing weird errors when we were testing our btrfs backports
before we had the incorrect level check fix.  These errors appeared to
be improper error handling, but error injection testing uncovered that
the errors were a result of corruption that occurred from improper error
handling during snapshot delete.

With snapshot delete if we encounter any errors during walk_down or
walk_up we'll simply return an error, we won't abort the transaction.
This is problematic because we will be dropping references for nodes and
leaves along the way, and if we fail in the middle we will leave the
file system corrupt because we don't know where we left off in the drop.

Fix this by making sure we abort if we hit any errors during the walk
down or walk up operations, as we have no idea what operations could
have been left half done at this point.

Signed-off-by: Josef Bacik <josef@toxicpanda.com>
Reviewed-by: David Sterba <dsterba@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
15 months agobtrfs: handle errors in walk_down_tree properly
Josef Bacik [Tue, 7 Feb 2023 16:57:24 +0000 (11:57 -0500)]
btrfs: handle errors in walk_down_tree properly

We can get errors in walk_down_proc as we try and lookup extent info for
the snapshot dropping to act on.  However if we get an error we simply
return 1 which indicates we're done with walking down, which will lead
us to improperly continue with the snapshot drop with the incorrect
information.  Instead break if we get any error from walk_down_proc or
do_walk_down, and handle the case of ret == 1 by returning 0, otherwise
return the ret value that we have.

Signed-off-by: Josef Bacik <josef@toxicpanda.com>
Reviewed-by: David Sterba <dsterba@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
15 months agobtrfs: drop root refs properly when orphan cleanup fails
Josef Bacik [Tue, 7 Feb 2023 16:57:23 +0000 (11:57 -0500)]
btrfs: drop root refs properly when orphan cleanup fails

When we mount the file system we do something like this:

while (1) {
lookup fs roots;

for (i = 0; i < num_roots; i++) {
ret = btrfs_orphan_cleanup(roots[i]);
if (ret)
break;
btrfs_put_root(roots[i]);
}
}

for (; i < num_roots; i++)
btrfs_put_root(roots[i]);

As you can see if we break in that inner loop we just go back to the
outer loop and lose the fact that we have to drop references on the
remaining roots we looked up.  Fix this by making an out label and
jumping to that on error so we don't leak a reference to the roots we
looked up.

Reviewed-by: Johannes Thumshirn <johannes.thumshirn@wdc.com>
Signed-off-by: Josef Bacik <josef@toxicpanda.com>
Reviewed-by: David Sterba <dsterba@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
15 months agobtrfs: add missing iputs on orphan cleanup failure
Josef Bacik [Tue, 7 Feb 2023 16:57:22 +0000 (11:57 -0500)]
btrfs: add missing iputs on orphan cleanup failure

We missed a couple of iput()s in the orphan cleanup failure paths, add
them so we don't get refcount errors. The iput needs to be done in the
check and not under a common label due to the way the code is
structured.

Signed-off-by: Josef Bacik <josef@toxicpanda.com>
Reviewed-by: David Sterba <dsterba@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
15 months agobtrfs: handle errors from btrfs_read_node_slot in split
Josef Bacik [Tue, 7 Feb 2023 16:57:21 +0000 (11:57 -0500)]
btrfs: handle errors from btrfs_read_node_slot in split

While investigating a problem with error injection I tripped over
curious behavior in the node/leaf splitting code.  If we get an EIO when
trying to read either the left or right leaf/node for splitting we'll
simply treat the node as if it were full and continue on.  The end
result of this isn't too bad, we simply end up allocating a block when
we may have pushed items into the adjacent blocks.

However this does essentially allow us to continue to modify a file
system that we've gotten errors on, either from a bad disk or csum
mismatch or other corruption.  This isn't particularly safe, so instead
handle these btrfs_read_node_slot() usages differently.  We allow you to
pass in any slot, the idea being that we save some code if the slot
number is outside of the range of the parent.  This means we treat all
errors the same, when in reality we only want to ignore -ENOENT.

Fix this by changing how we call btrfs_read_node_slot(), which is to
only call it for slots we know are valid.  This way if we get an error
back from reading the block we can properly pass the error up the chain.
This was validated with the error injection testing I was doing.

Signed-off-by: Josef Bacik <josef@toxicpanda.com>
Reviewed-by: David Sterba <dsterba@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
15 months agobtrfs: replace BUG_ON with ASSERT in btrfs_read_node_slot
Josef Bacik [Tue, 7 Feb 2023 16:57:20 +0000 (11:57 -0500)]
btrfs: replace BUG_ON with ASSERT in btrfs_read_node_slot

In btrfs_read_node_slot() we have a BUG_ON() that can be converted to an
ASSERT(), it's from an extent buffer and the level is validated at the
time it's read from disk.

Reviewed-by: Johannes Thumshirn <johannes.thumshirn@wdc.com>
Signed-off-by: Josef Bacik <josef@toxicpanda.com>
Reviewed-by: David Sterba <dsterba@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
15 months agobtrfs: use btrfs_handle_fs_error in btrfs_fill_super
Josef Bacik [Tue, 7 Feb 2023 16:57:19 +0000 (11:57 -0500)]
btrfs: use btrfs_handle_fs_error in btrfs_fill_super

While trying to track down a lost EIO problem I hit the following
assertion while doing my error injection testing

  BTRFS warning (device nvme1n1): transaction 1609 (with 180224 dirty metadata bytes) is not committed
  assertion failed: !found, in fs/btrfs/disk-io.c:4456
  ------------[ cut here ]------------
  kernel BUG at fs/btrfs/messages.h:169!
  invalid opcode: 0000 [#1] PREEMPT SMP NOPTI
  CPU: 0 PID: 1445 Comm: mount Tainted: G        W          6.2.0-rc5+ #3
  Hardware name: QEMU Standard PC (i440FX + PIIX, 1996), BIOS 1.16.1-2.fc37 04/01/2014
  RIP: 0010:btrfs_assertfail.constprop.0+0x18/0x1a
  RSP: 0018:ffffb95fc3b0bc68 EFLAGS: 00010286
  RAX: 0000000000000034 RBX: ffff9941c2ac2000 RCX: 0000000000000000
  RDX: 0000000000000001 RSI: ffffffffb6741f7d RDI: 00000000ffffffff
  RBP: ffff9941c2ac2428 R08: 0000000000000000 R09: ffffb95fc3b0bb38
  R10: 0000000000000003 R11: ffffffffb71438a8 R12: ffff9941c2ac2428
  R13: ffff9941c2ac2450 R14: ffff9941c2ac2450 R15: 000000000002c000
  FS:  00007fcea2d07800(0000) GS:ffff9941fbc00000(0000) knlGS:0000000000000000
  CS:  0010 DS: 0000 ES: 0000 CR0: 0000000080050033
  CR2: 00007f00cc7c83a8 CR3: 000000010c686000 CR4: 0000000000350ef0
  Call Trace:
   <TASK>
   close_ctree+0x426/0x48f
   btrfs_mount_root.cold+0x7e/0xee
   ? legacy_parse_param+0x2b/0x220
   legacy_get_tree+0x2b/0x50
   vfs_get_tree+0x29/0xc0
   vfs_kern_mount.part.0+0x73/0xb0
   btrfs_mount+0x11d/0x3d0
   ? legacy_parse_param+0x2b/0x220
   legacy_get_tree+0x2b/0x50
   vfs_get_tree+0x29/0xc0
   path_mount+0x438/0xa40
   __x64_sys_mount+0xe9/0x130
   do_syscall_64+0x3e/0x90
   entry_SYSCALL_64_after_hwframe+0x72/0xdc

This is because the error injection did an EIO for the root inode lookup
and we simply jumped to closing the ctree.  However because we didn't
mark the file system as having an error we skipped all of the broken
transaction cleanup stuff, and thus triggered this ASSERT().  Fix this
by calling btrfs_handle_fs_error() in this case so we have the error set
on the file system.

Reviewed-by: Johannes Thumshirn <johannes.thumshirn@wdc.com>
Signed-off-by: Josef Bacik <josef@toxicpanda.com>
Reviewed-by: David Sterba <dsterba@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
15 months agoLinux 6.3-rc7
Linus Torvalds [Sun, 16 Apr 2023 22:23:53 +0000 (15:23 -0700)]
Linux 6.3-rc7

15 months agoMerge tag 'sched_urgent_for_v6.3_rc7' of git://git.kernel.org/pub/scm/linux/kernel...
Linus Torvalds [Sun, 16 Apr 2023 17:33:43 +0000 (10:33 -0700)]
Merge tag 'sched_urgent_for_v6.3_rc7' of git://git./linux/kernel/git/tip/tip

Pull scheduler fix from Borislav Petkov:

 - Do not pull tasks to the local scheduling group if its average load
   is higher than the average system load

* tag 'sched_urgent_for_v6.3_rc7' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip:
  sched/fair: Fix imbalance overflow

15 months agoMerge tag 'x86_urgent_for_v6.3_rc7' of git://git.kernel.org/pub/scm/linux/kernel...
Linus Torvalds [Sun, 16 Apr 2023 17:28:29 +0000 (10:28 -0700)]
Merge tag 'x86_urgent_for_v6.3_rc7' of git://git./linux/kernel/git/tip/tip

Pull x86 fix from Borislav Petkov:

 - Drop __init annotation from two rtc functions which get called after
   boot is done, in order to prevent a crash

* tag 'x86_urgent_for_v6.3_rc7' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip:
  x86/rtc: Remove __init for runtime functions

15 months agoMerge tag 'powerpc-6.3-5' of git://git.kernel.org/pub/scm/linux/kernel/git/powerpc...
Linus Torvalds [Sun, 16 Apr 2023 16:55:18 +0000 (09:55 -0700)]
Merge tag 'powerpc-6.3-5' of git://git./linux/kernel/git/powerpc/linux

Pull powerpc fix from Michael Ellerman:

 - A fix for NUMA distance handling in the pseries SCM (pmem) driver.

   Thanks to Aneesh Kumar K.V.

* tag 'powerpc-6.3-5' of git://git.kernel.org/pub/scm/linux/kernel/git/powerpc/linux:
  powerpc/papr_scm: Update the NUMA distance table for the target node

15 months agoMerge tag 'kbuild-fixes-v6.3-3' of git://git.kernel.org/pub/scm/linux/kernel/git...
Linus Torvalds [Sun, 16 Apr 2023 16:46:32 +0000 (09:46 -0700)]
Merge tag 'kbuild-fixes-v6.3-3' of git://git./linux/kernel/git/masahiroy/linux-kbuild

Pull Kbuild fixes from Masahiro Yamada:

 - Drop debug info from purgatory objects again

 - Document that kernel.org provides prebuilt LLVM toolchains

 - Give up handling untracked files for source package builds

 - Avoid creating corrupted cpio when KBUILD_BUILD_TIMESTAMP is given
   with a pre-epoch data.

 - Change panic_show_mem() to a macro to handle variable-length argument

 - Compress tarballs on-the-fly again

* tag 'kbuild-fixes-v6.3-3' of git://git.kernel.org/pub/scm/linux/kernel/git/masahiroy/linux-kbuild:
  kbuild: do not create intermediate *.tar for tar packages
  kbuild: do not create intermediate *.tar for source tarballs
  kbuild: merge cmd_archive_linux and cmd_archive_perf
  init/initramfs: Fix argument forwarding to panic() in panic_show_mem()
  initramfs: Check negative timestamp to prevent broken cpio archive
  kbuild: give up untracked files for source package builds
  Documentation/llvm: Add a note about prebuilt kernel.org toolchains
  purgatory: fix disabling debug info

15 months agoMerge tag '6.3-rc6-ksmbd-server-fix' of git://git.samba.org/ksmbd
Linus Torvalds [Sun, 16 Apr 2023 16:39:55 +0000 (09:39 -0700)]
Merge tag '6.3-rc6-ksmbd-server-fix' of git://git.samba.org/ksmbd

Pull ksmbd server fix from Steve French:
 "smb311 server preauth integrity negotiate context parsing fix (check
  for out of bounds access)"

* tag '6.3-rc6-ksmbd-server-fix' of git://git.samba.org/ksmbd:
  ksmbd: avoid out of bounds access in decode_preauth_ctxt()

15 months agokbuild: do not create intermediate *.tar for tar packages
Masahiro Yamada [Fri, 7 Apr 2023 10:16:29 +0000 (19:16 +0900)]
kbuild: do not create intermediate *.tar for tar packages

Commit 05e96e96a315 ("kbuild: use git-archive for source package
creation") split the compression as a separate step to factor out
the common build rules.

With the previous commit, we got back to the situation where source
tarballs are compressed on-the-fly.
There is no reason to keep the separate compression rules.

Generate the comressed tar packages directly.

Signed-off-by: Masahiro Yamada <masahiroy@kernel.org>
Reviewed-by: Nathan Chancellor <nathan@kernel.org>
15 months agokbuild: do not create intermediate *.tar for source tarballs
Masahiro Yamada [Fri, 7 Apr 2023 10:16:28 +0000 (19:16 +0900)]
kbuild: do not create intermediate *.tar for source tarballs

Since commit 05e96e96a315 ("kbuild: use git-archive for source package
creation"), a source tarball is created in two steps; create *.tar file
then compress it. I split the compression as a separate rule because I
just thought 'git archive' supported only gzip.

For other compression algorithms, I could pipe the two commands:

  $ git archive HEAD | xz > linux.tar.xz

I read git-archive(1) carefully, and I realized GIT had provided a
more elegant way:

  $ git -c tar.tar.xz.command=xz archive -o linux.tar.xz HEAD

This commit uses 'tar.tar.*.command' configuration to specify the
compression backend so we can compress a source tarball on-the-fly.

GIT commit 767cf4579f0e ("archive: implement configurable tar filters")
is more than a decade old, so it should be available on almost all build
environments.

Signed-off-by: Masahiro Yamada <masahiroy@kernel.org>
Reviewed-by: Nathan Chancellor <nathan@kernel.org>
15 months agokbuild: merge cmd_archive_linux and cmd_archive_perf
Masahiro Yamada [Fri, 7 Apr 2023 10:16:27 +0000 (19:16 +0900)]
kbuild: merge cmd_archive_linux and cmd_archive_perf

The two commands, cmd_archive_linux and cmd_archive_perf, are similar.
Merge them to make it easier to add more changes to the git-archive
command.

Signed-off-by: Masahiro Yamada <masahiroy@kernel.org>
Reviewed-by: Nathan Chancellor <nathan@kernel.org>
15 months agoinit/initramfs: Fix argument forwarding to panic() in panic_show_mem()
Benjamin Gray [Mon, 20 Mar 2023 23:05:34 +0000 (10:05 +1100)]
init/initramfs: Fix argument forwarding to panic() in panic_show_mem()

Forwarding variadic argument lists can't be done by passing a va_list
to a function with signature foo(...) (as panic() has). It ends up
interpreting the va_list itself as a single argument instead of
iterating it. printf() happily accepts it of course, leading to corrupt
output.

Convert panic_show_mem() to a macro to allow forwarding the arguments.
The function is trivial enough that it's easier than trying to introduce
a vpanic() variant.

Signed-off-by: Benjamin Gray <bgray@linux.ibm.com>
Reviewed-by: Andrew Donnellan <ajd@linux.ibm.com>
Signed-off-by: Masahiro Yamada <masahiroy@kernel.org>
15 months agoinitramfs: Check negative timestamp to prevent broken cpio archive
Benjamin Gray [Mon, 20 Mar 2023 04:08:38 +0000 (15:08 +1100)]
initramfs: Check negative timestamp to prevent broken cpio archive

Similar to commit 4c9d410f32b3 ("initramfs: Check timestamp to prevent
broken cpio archive"), except asserts that the timestamp is
non-negative. This can happen when the KBUILD_BUILD_TIMESTAMP is a value
before UNIX epoch, which may be set when making reproducible builds that
don't want to look like they use a valid date.

While support for dates before 1970 might not be supported, this is more
about preventing undetected CPIO corruption. The printf's use a minimum
length format specifier, and will happily make the field longer than 8
characters if they need to.

Signed-off-by: Benjamin Gray <bgray@linux.ibm.com>
Reviewed-by: Andrew Donnellan <ajd@linux.ibm.com>
Tested-by: Andrew Donnellan <ajd@linux.ibm.com>
Signed-off-by: Masahiro Yamada <masahiroy@kernel.org>
15 months agoMerge tag '6.3-rc6-smb311-client-negcontext-fix' of git://git.samba.org/sfrench/cifs-2.6
Linus Torvalds [Sun, 16 Apr 2023 01:37:51 +0000 (18:37 -0700)]
Merge tag '6.3-rc6-smb311-client-negcontext-fix' of git://git.samba.org/sfrench/cifs-2.6

Pull cifs fix from Steve French:
 "Small client fix for better checking for smb311 negotiate context
  overflows, also marked for stable"

* tag '6.3-rc6-smb311-client-negcontext-fix' of git://git.samba.org/sfrench/cifs-2.6:
  cifs: fix negotiate context parsing

15 months agoMerge tag 'ubifs-for-linus-6.3-rc7' of git://git.kernel.org/pub/scm/linux/kernel...
Linus Torvalds [Sat, 15 Apr 2023 23:55:09 +0000 (16:55 -0700)]
Merge tag 'ubifs-for-linus-6.3-rc7' of git://git./linux/kernel/git/rw/ubifs

Pull UBI fixes from Richard Weinberger:

 - Fix failure to attach when vid_hdr offset equals the (sub)page size

 - Fix for a deadlock in UBI's worker thread

* tag 'ubifs-for-linus-6.3-rc7' of git://git.kernel.org/pub/scm/linux/kernel/git/rw/ubifs:
  ubi: Fix failure attaching when vid_hdr offset equals to (sub)page size
  ubi: Fix deadlock caused by recursively holding work_sem

15 months agocifs: fix negotiate context parsing
David Disseldorp [Thu, 6 Apr 2023 22:34:11 +0000 (00:34 +0200)]
cifs: fix negotiate context parsing

smb311_decode_neg_context() doesn't properly check against SMB packet
boundaries prior to accessing individual negotiate context entries. This
is due to the length check omitting the eight byte smb2_neg_context
header, as well as incorrect decrementing of len_of_ctxts.

Fixes: 5100d8a3fe03 ("SMB311: Improve checking of negotiate security contexts")
Reported-by: Volker Lendecke <vl@samba.org>
Reviewed-by: Paulo Alcantara (SUSE) <pc@manguebit.com>
Signed-off-by: David Disseldorp <ddiss@suse.de>
Signed-off-by: Steve French <stfrench@microsoft.com>
15 months agoMerge tag 'i2c-for-6.3-rc7' of git://git.kernel.org/pub/scm/linux/kernel/git/wsa...
Linus Torvalds [Sat, 15 Apr 2023 18:06:49 +0000 (11:06 -0700)]
Merge tag 'i2c-for-6.3-rc7' of git://git./linux/kernel/git/wsa/linux

Pull i2c fixes from Wolfram Sang:
 "Just two driver fixes"

* tag 'i2c-for-6.3-rc7' of git://git.kernel.org/pub/scm/linux/kernel/git/wsa/linux:
  i2c: ocores: generate stop condition after timeout in polling mode
  i2c: mchp-pci1xxxx: Update Timing registers

15 months agoMerge tag 'scsi-fixes' of git://git.kernel.org/pub/scm/linux/kernel/git/jejb/scsi
Linus Torvalds [Sat, 15 Apr 2023 17:49:47 +0000 (10:49 -0700)]
Merge tag 'scsi-fixes' of git://git./linux/kernel/git/jejb/scsi

Pull SCSI fix from James Bottomley:
 "One small fix to SCSI Enclosure Services to fix a regression caused by
  another recent fix"

* tag 'scsi-fixes' of git://git.kernel.org/pub/scm/linux/kernel/git/jejb/scsi:
  scsi: ses: Handle enclosure with just a primary component gracefully

15 months agoMerge tag 'block-6.3-2023-04-14' of git://git.kernel.dk/linux
Linus Torvalds [Sat, 15 Apr 2023 17:39:18 +0000 (10:39 -0700)]
Merge tag 'block-6.3-2023-04-14' of git://git.kernel.dk/linux

Pull block fix from Jens Axboe:
 "A single NVMe quirk entry addition"

* tag 'block-6.3-2023-04-14' of git://git.kernel.dk/linux:
  nvme-pci: add NVME_QUIRK_BOGUS_NID for T-FORCE Z330 SSD

15 months agoMerge tag 'io_uring-6.3-2023-04-14' of git://git.kernel.dk/linux
Linus Torvalds [Sat, 15 Apr 2023 17:29:53 +0000 (10:29 -0700)]
Merge tag 'io_uring-6.3-2023-04-14' of git://git.kernel.dk/linux

Pull io_uring fix from Jens Axboe:
 "Just a small tweak to when task_work needs redirection, marked for
  stable as well"

* tag 'io_uring-6.3-2023-04-14' of git://git.kernel.dk/linux:
  io_uring: complete request via task work in case of DEFER_TASKRUN

15 months agoMerge tag 'riscv-for-linus-6.3-rc7' of git://git.kernel.org/pub/scm/linux/kernel...
Linus Torvalds [Fri, 14 Apr 2023 17:44:48 +0000 (10:44 -0700)]
Merge tag 'riscv-for-linus-6.3-rc7' of git://git./linux/kernel/git/riscv/linux

Pull RISC-V fixes from Palmer Dabbelt:

 - A fix for a missing fence when generating the NOMMU sigreturn
   trampoline

 - A set of fixes for early DTB handling of reserved memory nodes

* tag 'riscv-for-linus-6.3-rc7' of git://git.kernel.org/pub/scm/linux/kernel/git/riscv/linux:
  riscv: No need to relocate the dtb as it lies in the fixmap region
  riscv: Do not set initial_boot_params to the linear address of the dtb
  riscv: Move early dtb mapping into the fixmap region
  riscv: add icache flush for nommu sigreturn trampoline

15 months agoMerge tag 'acpi-6.3-rc7' of git://git.kernel.org/pub/scm/linux/kernel/git/rafael...
Linus Torvalds [Fri, 14 Apr 2023 17:37:07 +0000 (10:37 -0700)]
Merge tag 'acpi-6.3-rc7' of git://git./linux/kernel/git/rafael/linux-pm

Pull ACPI fixes from Rafael Wysocki:
 "These add two ACPI-related quirks:

   - Add a quirk to force StorageD3Enable on AMD Picasso systems (Mario
     Limonciello)

   - Add an ACPI IRQ override quirk for ASUS ExpertBook B1502CBA (Paul
     Menzel)"

* tag 'acpi-6.3-rc7' of git://git.kernel.org/pub/scm/linux/kernel/git/rafael/linux-pm:
  ACPI: resource: Skip IRQ override on ASUS ExpertBook B1502CBA
  ACPI: x86: utils: Add Picasso to the list for forcing StorageD3Enable

15 months agoMerge tag 'pm-6.3-rc7' of git://git.kernel.org/pub/scm/linux/kernel/git/rafael/linux-pm
Linus Torvalds [Fri, 14 Apr 2023 17:25:30 +0000 (10:25 -0700)]
Merge tag 'pm-6.3-rc7' of git://git./linux/kernel/git/rafael/linux-pm

Pull power management fix from Rafael Wysocki:
 "Make the amd-pstate cpufreq driver take all of the possible
  combinations of the 'old' and 'new' status values correctly while
  changing the operation mode via sysfs (Wyes Karny)"

* tag 'pm-6.3-rc7' of git://git.kernel.org/pub/scm/linux/kernel/git/rafael/linux-pm:
  amd-pstate: Fix amd_pstate mode switch

15 months agoMerge tag 'thermal-6.3-rc7' of git://git.kernel.org/pub/scm/linux/kernel/git/rafael...
Linus Torvalds [Fri, 14 Apr 2023 17:19:18 +0000 (10:19 -0700)]
Merge tag 'thermal-6.3-rc7' of git://git./linux/kernel/git/rafael/linux-pm

Pull thermal control fix from Rafael Wysocki:
 "Modify the Intel thermal throttling code to avoid updating unsupported
  status clearing mask bits which causes the kernel to complain about
  unchecked MSR access (Srinivas Pandruvada)"

* tag 'thermal-6.3-rc7' of git://git.kernel.org/pub/scm/linux/kernel/git/rafael/linux-pm:
  thermal: intel: Avoid updating unsupported THERM_STATUS_CLEAR mask bits

15 months agoMerge tag 'sound-6.3-rc7' of git://git.kernel.org/pub/scm/linux/kernel/git/tiwai...
Linus Torvalds [Fri, 14 Apr 2023 17:13:54 +0000 (10:13 -0700)]
Merge tag 'sound-6.3-rc7' of git://git./linux/kernel/git/tiwai/sound

Pull sound fixes from Takashi Iwai:
 "A collection of small fixes.

  At this time, quite a few fixes for the old PCI drivers are found.
  Although they are not regression fixes, I took these as they are
  materials for stable kernels.

  In addition, a couple of regression fixes and another couple of
  HD-audio quirks are included"

* tag 'sound-6.3-rc7' of git://git.kernel.org/pub/scm/linux/kernel/git/tiwai/sound:
  ALSA: hda/hdmi: disable KAE for Intel DG2
  ALSA: hda/realtek: Add quirks for Lenovo Z13/Z16 Gen2
  ALSA: hda: patch_realtek: add quirk for Asus N7601ZM
  ALSA: firewire-tascam: add missing unwind goto in snd_tscm_stream_start_duplex()
  ALSA: emu10k1: don't create old pass-through playback device on Audigy
  ALSA: emu10k1: fix capture interrupt handler unlinking
  ALSA: hda/sigmatel: fix S/PDIF out on Intel D*45* motherboards
  ALSA: hda/sigmatel: add pin overrides for Intel DP45SG motherboard
  ALSA: i2c/cs8427: fix iec958 mixer control deactivation

15 months agoMerge tag 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/rdma/rdma
Linus Torvalds [Fri, 14 Apr 2023 17:06:50 +0000 (10:06 -0700)]
Merge tag 'for-linus' of git://git./linux/kernel/git/rdma/rdma

Pull rdma fixes from Jason Gunthorpe:
 "We had a fairly slow cycle on the rc side this time, here are the
  accumulated fixes, mostly in drivers:

   - irdma should not generate extra completions during flushing

   - Fix several memory leaks

   - Do not get confused in irdma's iwarp mode if IPv6 is present

   - Correct a link speed calculation in mlx5

   - Increase the EQ/WQ limits on erdma as they are too small for big
     applications

   - Use the right math for erdma's inline mtt feature

   - Make erdma probing more robust to boot time ordering differences

   - Fix a KMSAN crash in CMA due to uninitialized qkey"

* tag 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/rdma/rdma:
  RDMA/core: Fix GID entry ref leak when create_ah fails
  RDMA/cma: Allow UD qp_type to join multicast only
  RDMA/erdma: Defer probing if netdevice can not be found
  RDMA/erdma: Inline mtt entries into WQE if supported
  RDMA/erdma: Update default EQ depth to 4096 and max_send_wr to 8192
  RDMA/erdma: Fix some typos
  IB/mlx5: Add support for 400G_8X lane speed
  RDMA/irdma: Add ipv4 check to irdma_find_listener()
  RDMA/irdma: Increase iWARP CM default rexmit count
  RDMA/irdma: Fix memory leak of PBLE objects
  RDMA/irdma: Do not generate SW completions for NOPs

15 months agoMerge branch 'acpi-x86'
Rafael J. Wysocki [Fri, 14 Apr 2023 13:15:32 +0000 (15:15 +0200)]
Merge branch 'acpi-x86'

Merge a quirk to force StorageD3Enable on AMD Picasso systems (Mario
Limonciello).

* acpi-x86:
  ACPI: x86: utils: Add Picasso to the list for forcing StorageD3Enable

15 months agoio_uring: complete request via task work in case of DEFER_TASKRUN
Ming Lei [Fri, 14 Apr 2023 07:53:13 +0000 (15:53 +0800)]
io_uring: complete request via task work in case of DEFER_TASKRUN

So far io_req_complete_post() only covers DEFER_TASKRUN by completing
request via task work when the request is completed from IOWQ.

However, uring command could be completed from any context, and if io
uring is setup with DEFER_TASKRUN, the command is required to be
completed from current context, otherwise wait on IORING_ENTER_GETEVENTS
can't be wakeup, and may hang forever.

The issue can be observed on removing ublk device, but turns out it is
one generic issue for uring command & DEFER_TASKRUN, so solve it in
io_uring core code.

Fixes: e6aeb2721d3b ("io_uring: complete all requests in task context")
Cc: stable@vger.kernel.org
Link: https://lore.kernel.org/linux-block/b3fc9991-4c53-9218-a8cc-5b4dd3952108@kernel.dk/
Reported-by: Jens Axboe <axboe@kernel.dk>
Cc: Kanchan Joshi <joshi.k@samsung.com>
Signed-off-by: Ming Lei <ming.lei@redhat.com>
Signed-off-by: Jens Axboe <axboe@kernel.dk>
15 months agoMerge branch 'nvme-6.3' of git://git.infradead.org/nvme into block-6.3
Jens Axboe [Fri, 14 Apr 2023 12:29:00 +0000 (06:29 -0600)]
Merge branch 'nvme-6.3' of git://git.infradead.org/nvme into block-6.3

Pull NVMe fix from Christoph.

* 'nvme-6.3' of git://git.infradead.org/nvme:
  nvme-pci: add NVME_QUIRK_BOGUS_NID for T-FORCE Z330 SSD

15 months agoALSA: hda/hdmi: disable KAE for Intel DG2
Kai Vehmanen [Thu, 13 Apr 2023 19:11:53 +0000 (22:11 +0300)]
ALSA: hda/hdmi: disable KAE for Intel DG2

Use of keep-alive (KAE) has resulted in loss of audio on some A750/770
cards as the transition from keep-alive to stream playback is not
working as expected. As there is limited benefit of the new KAE mode
on discrete cards, revert back to older silent-stream implementation
on these systems.

Cc: stable@vger.kernel.org
Fixes: 15175a4f2bbb ("ALSA: hda/hdmi: add keep-alive support for ADL-P and DG2")
Link: https://gitlab.freedesktop.org/drm/intel/-/issues/8307
Signed-off-by: Kai Vehmanen <kai.vehmanen@linux.intel.com>
Link: https://lore.kernel.org/r/20230413191153.3692049-1-kai.vehmanen@linux.intel.com
Signed-off-by: Takashi Iwai <tiwai@suse.de>
15 months agonvme-pci: add NVME_QUIRK_BOGUS_NID for T-FORCE Z330 SSD
Duy Truong [Fri, 14 Apr 2023 00:55:48 +0000 (17:55 -0700)]
nvme-pci: add NVME_QUIRK_BOGUS_NID for T-FORCE Z330 SSD

Added a quirk to fix the TeamGroup T-Force Cardea Zero Z330 SSDs reporting
duplicate NGUIDs.

Signed-off-by: Duy Truong <dory@dory.moe>
Cc: stable@vger.kernel.org
Signed-off-by: Christoph Hellwig <hch@lst.de>
15 months agoriscv: No need to relocate the dtb as it lies in the fixmap region
Alexandre Ghiti [Wed, 29 Mar 2023 08:19:32 +0000 (10:19 +0200)]
riscv: No need to relocate the dtb as it lies in the fixmap region

We used to access the dtb via its linear mapping address but now that the
dtb early mapping was moved in the fixmap region, we can keep using this
address since it is present in swapper_pg_dir, and remove the dtb
relocation.

Note that the relocation was wrong anyway since early_memremap() is
restricted to 256K whereas the maximum fdt size is 2MB.

Signed-off-by: Alexandre Ghiti <alexghiti@rivosinc.com>
Reviewed-by: Conor Dooley <conor.dooley@microchip.com>
Tested-by: Conor Dooley <conor.dooley@microchip.com>
Link: https://lore.kernel.org/r/20230329081932.79831-4-alexghiti@rivosinc.com
Cc: stable@vger.kernel.org
Signed-off-by: Palmer Dabbelt <palmer@rivosinc.com>
15 months agoriscv: Do not set initial_boot_params to the linear address of the dtb
Alexandre Ghiti [Wed, 29 Mar 2023 08:19:31 +0000 (10:19 +0200)]
riscv: Do not set initial_boot_params to the linear address of the dtb

early_init_dt_verify() is already called in parse_dtb() and since the dtb
address does not change anymore (it is now in the fixmap region), no need
to reset initial_boot_params by calling early_init_dt_verify() again.

Signed-off-by: Alexandre Ghiti <alexghiti@rivosinc.com>
Link: https://lore.kernel.org/r/20230329081932.79831-3-alexghiti@rivosinc.com
Cc: stable@vger.kernel.org
Signed-off-by: Palmer Dabbelt <palmer@rivosinc.com>
15 months agoriscv: Move early dtb mapping into the fixmap region
Alexandre Ghiti [Wed, 29 Mar 2023 08:19:30 +0000 (10:19 +0200)]
riscv: Move early dtb mapping into the fixmap region

riscv establishes 2 virtual mappings:

- early_pg_dir maps the kernel which allows to discover the system
  memory
- swapper_pg_dir installs the final mapping (linear mapping included)

We used to map the dtb in early_pg_dir using DTB_EARLY_BASE_VA, and this
mapping was not carried over in swapper_pg_dir. It happens that
early_init_fdt_scan_reserved_mem() must be called before swapper_pg_dir is
setup otherwise we could allocate reserved memory defined in the dtb.
And this function initializes reserved_mem variable with addresses that
lie in the early_pg_dir dtb mapping: when those addresses are reused
with swapper_pg_dir, this mapping does not exist and then we trap.

The previous "fix" was incorrect as early_init_fdt_scan_reserved_mem()
must be called before swapper_pg_dir is set up otherwise we could
allocate in reserved memory defined in the dtb.

So move the dtb mapping in the fixmap region which is established in
early_pg_dir and handed over to swapper_pg_dir.

Fixes: 922b0375fc93 ("riscv: Fix memblock reservation for device tree blob")
Fixes: 8f3a2b4a96dc ("RISC-V: Move DT mapping outof fixmap")
Fixes: 50e63dd8ed92 ("riscv: fix reserved memory setup")
Reported-by: Conor Dooley <conor.dooley@microchip.com>
Link: https://lore.kernel.org/all/f8e67f82-103d-156c-deb0-d6d6e2756f5e@microchip.com/
Signed-off-by: Alexandre Ghiti <alexghiti@rivosinc.com>
Reviewed-by: Conor Dooley <conor.dooley@microchip.com>
Tested-by: Conor Dooley <conor.dooley@microchip.com>
Link: https://lore.kernel.org/r/20230329081932.79831-2-alexghiti@rivosinc.com
Cc: stable@vger.kernel.org
Signed-off-by: Palmer Dabbelt <palmer@rivosinc.com>
15 months agoMerge tag 'cgroup-for-6.3-rc6-fixes' of git://git.kernel.org/pub/scm/linux/kernel...
Linus Torvalds [Thu, 13 Apr 2023 23:28:33 +0000 (16:28 -0700)]
Merge tag 'cgroup-for-6.3-rc6-fixes' of git://git./linux/kernel/git/tj/cgroup

Pull cgroup fixes from Tejun Heo:
 "This is a relatively big pull request this late in the cycle but the
  major contributor is the cpuset bug which is rather significant:

   - Fix several cpuset bugs including one where it wasn't applying the
     target cgroup when tasks are created with CLONE_INTO_CGROUP

  With a few smaller fixes:

   - Fix inversed locking order in cgroup1 freezer implementation

   - Fix garbage cpu.stat::core_sched.forceidle_usec reporting in the
     root cgroup"

* tag 'cgroup-for-6.3-rc6-fixes' of git://git.kernel.org/pub/scm/linux/kernel/git/tj/cgroup:
  cgroup/cpuset: Make cpuset_attach_task() skip subpartitions CPUs for top_cpuset
  cgroup/cpuset: Add cpuset_can_fork() and cpuset_cancel_fork() methods
  cgroup/cpuset: Make cpuset_fork() handle CLONE_INTO_CGROUP properly
  cgroup/cpuset: Wake up cpuset_attach_wq tasks in cpuset_cancel_attach()
  cgroup,freezer: hold cpu_hotplug_lock before freezer_mutex
  cgroup/cpuset: Fix partition root's cpuset.cpus update bug
  cgroup: fix display of forceidle time at root

15 months agoMerge tag 'clk-fixes-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git...
Linus Torvalds [Thu, 13 Apr 2023 23:16:33 +0000 (16:16 -0700)]
Merge tag 'clk-fixes-for-linus' of git://git./linux/kernel/git/clk/linux

Pull clk fixes from Stephen Boyd:
 "A few more clk driver fixes:

   - Set the max_register member of the spreadtrum regmap so that reads
     don't go off the end of the I/O space

   - Avoid a clk parent error in the i.MX imx6ul driver when the
     selector is unknown

   - Fix an oops due to REGCACHE_NONE usage by the Renesas 9-series
     driver"

* tag 'clk-fixes-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/clk/linux:
  clk: rs9: Fix suspend/resume
  clk: imx6ul: fix "failed to get parent" error
  clk: sprd: set max_register according to mapping range

15 months agoMerge tag 'net-6.3-rc7' of git://git.kernel.org/pub/scm/linux/kernel/git/netdev/net
Linus Torvalds [Thu, 13 Apr 2023 22:33:04 +0000 (15:33 -0700)]
Merge tag 'net-6.3-rc7' of git://git./linux/kernel/git/netdev/net

Pull networking fixes from Jakub Kicinski:
 "Including fixes from bpf, and bluetooth.

  Not all that quiet given spring celebrations, but "current" fixes are
  thinning out, which is encouraging. One outstanding regression in the
  mlx5 driver when using old FW, not blocking but we're pushing for a
  fix.

  Current release - new code bugs:

   - eth: enetc: workaround for unresponsive pMAC after receiving
     express traffic

  Previous releases - regressions:

   - rtnetlink: restore RTM_NEW/DELLINK notification behavior, keep the
     pid/seq fields 0 for backward compatibility

  Previous releases - always broken:

   - sctp: fix a potential overflow in sctp_ifwdtsn_skip

   - mptcp:
      - use mptcp_schedule_work instead of open-coding it and make the
        worker check stricter, to avoid scheduling work on closed
        sockets
      - fix NULL pointer dereference on fastopen early fallback

   - skbuff: fix memory corruption due to a race between skb coalescing
     and releasing clones confusing page_pool reference counting

   - bonding: fix neighbor solicitation validation on backup slaves

   - bpf: tcp: use sock_gen_put instead of sock_put in bpf_iter_tcp

   - bpf: arm64: fixed a BTI error on returning to patched function

   - openvswitch: fix race on port output leading to inf loop

   - sfp: initialize sfp->i2c_block_size at sfp allocation to avoid
     returning a different errno than expected

   - phy: nxp-c45-tja11xx: unregister PTP, purge queues on remove

   - Bluetooth: fix printing errors if LE Connection times out

   - Bluetooth: assorted UaF, deadlock and data race fixes

   - eth: macb: fix memory corruption in extended buffer descriptor mode

  Misc:

   - adjust the XDP Rx flow hash API to also include the protocol layers
     over which the hash was computed"

* tag 'net-6.3-rc7' of git://git.kernel.org/pub/scm/linux/kernel/git/netdev/net: (50 commits)
  selftests/bpf: Adjust bpf_xdp_metadata_rx_hash for new arg
  mlx4: bpf_xdp_metadata_rx_hash add xdp rss hash type
  veth: bpf_xdp_metadata_rx_hash add xdp rss hash type
  mlx5: bpf_xdp_metadata_rx_hash add xdp rss hash type
  xdp: rss hash types representation
  selftests/bpf: xdp_hw_metadata remove bpf_printk and add counters
  skbuff: Fix a race between coalescing and releasing SKBs
  net: macb: fix a memory corruption in extended buffer descriptor mode
  selftests: add the missing CONFIG_IP_SCTP in net config
  udp6: fix potential access to stale information
  selftests: openvswitch: adjust datapath NL message declaration
  selftests: mptcp: userspace pm: uniform verify events
  mptcp: fix NULL pointer dereference on fastopen early fallback
  mptcp: stricter state check in mptcp_worker
  mptcp: use mptcp_schedule_work instead of open-coding it
  net: enetc: workaround for unresponsive pMAC after receiving express traffic
  sctp: fix a potential overflow in sctp_ifwdtsn_skip
  net: qrtr: Fix an uninit variable access bug in qrtr_tx_resume()
  rtnetlink: Restore RTM_NEW/DELLINK notification behavior
  net: ti/cpsw: Add explicit platform_device.h and of_platform.h includes
  ...

15 months agoMerge tag 'devicetree-fixes-for-6.2-3' of git://git.kernel.org/pub/scm/linux/kernel...
Linus Torvalds [Thu, 13 Apr 2023 22:21:56 +0000 (15:21 -0700)]
Merge tag 'devicetree-fixes-for-6.2-3' of git://git./linux/kernel/git/robh/linux

Pull devicetree fixes from Rob Herring:

 - Fix interaction between fw_devlink and DT overlays causing devices to
   not be probed

 - Fix the compatible string for loongson,cpu-interrupt-controller

* tag 'devicetree-fixes-for-6.2-3' of git://git.kernel.org/pub/scm/linux/kernel/git/robh/linux:
  treewide: Fix probing of devices in DT overlays
  dt-bindings: interrupt-controller: loongarch: Fix mismatched compatible

15 months agoMerge tag 'pinctrl-v6.3-3' of git://git.kernel.org/pub/scm/linux/kernel/git/linusw...
Linus Torvalds [Thu, 13 Apr 2023 22:17:59 +0000 (15:17 -0700)]
Merge tag 'pinctrl-v6.3-3' of git://git./linux/kernel/git/linusw/linux-pinctrl

Pull pin control fix from Linus Walleij:
 "This is just a revert of the AMD fix, because the fix broke some
  laptops. We are working on a proper solution"

* tag 'pinctrl-v6.3-3' of git://git.kernel.org/pub/scm/linux/kernel/git/linusw/linux-pinctrl:
  Revert "pinctrl: amd: Disable and mask interrupts on resume"

15 months agoMerge tag 'drm-fixes-2023-04-13' of git://anongit.freedesktop.org/drm/drm
Linus Torvalds [Thu, 13 Apr 2023 21:58:55 +0000 (14:58 -0700)]
Merge tag 'drm-fixes-2023-04-13' of git://anongit.freedesktop.org/drm/drm

Pull drm fixes from Daniel Vetter:

 - two fbcon regressions

 - amdgpu: dp mst, smu13

 - i915: dual link dsi for tgl+

 - armada, nouveau, drm/sched, fbmem

* tag 'drm-fixes-2023-04-13' of git://anongit.freedesktop.org/drm/drm:
  fbcon: set_con2fb_map needs to set con2fb_map!
  fbcon: Fix error paths in set_con2fb_map
  drm/amd/pm: correct the pcie link state check for SMU13
  drm/amd/pm: correct SMU13.0.7 max shader clock reporting
  drm/amd/pm: correct SMU13.0.7 pstate profiling clock settings
  drm/amd/display: Pass the right info to drm_dp_remove_payload
  drm/armada: Fix a potential double free in an error handling path
  fbmem: Reject FB_ACTIVATE_KD_TEXT from userspace
  drm/nouveau/fb: add missing sysmen flush callbacks
  drm/i915/dsi: fix DSS CTL register offsets for TGL+
  drm/scheduler: Fix UAF race in drm_sched_entity_push_job()

15 months agoMerge tag 'for-netdev' of https://git.kernel.org/pub/scm/linux/kernel/git/bpf/bpf
Jakub Kicinski [Thu, 13 Apr 2023 20:04:44 +0000 (13:04 -0700)]
Merge tag 'for-netdev' of https://git./linux/kernel/git/bpf/bpf

Daniel Borkmann says:

====================
pull-request: bpf 2023-04-13

We've added 6 non-merge commits during the last 1 day(s) which contain
a total of 14 files changed, 205 insertions(+), 38 deletions(-).

The main changes are:

1) One late straggler fix on the XDP hints side which fixes
   bpf_xdp_metadata_rx_hash kfunc API before the release goes out
   in order to provide information on the RSS hash type,
   from Jesper Dangaard Brouer.

* tag 'for-netdev' of https://git.kernel.org/pub/scm/linux/kernel/git/bpf/bpf:
  selftests/bpf: Adjust bpf_xdp_metadata_rx_hash for new arg
  mlx4: bpf_xdp_metadata_rx_hash add xdp rss hash type
  veth: bpf_xdp_metadata_rx_hash add xdp rss hash type
  mlx5: bpf_xdp_metadata_rx_hash add xdp rss hash type
  xdp: rss hash types representation
  selftests/bpf: xdp_hw_metadata remove bpf_printk and add counters
====================

Link: https://lore.kernel.org/r/20230413192939.10202-1-daniel@iogearbox.net
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
15 months agoksmbd: avoid out of bounds access in decode_preauth_ctxt()
David Disseldorp [Thu, 13 Apr 2023 14:49:57 +0000 (23:49 +0900)]
ksmbd: avoid out of bounds access in decode_preauth_ctxt()

Confirm that the accessed pneg_ctxt->HashAlgorithms address sits within
the SMB request boundary; deassemble_neg_contexts() only checks that the
eight byte smb2_neg_context header + (client controlled) DataLength are
within the packet boundary, which is insufficient.

Checking for sizeof(struct smb2_preauth_neg_context) is overkill given
that the type currently assumes SMB311_SALT_SIZE bytes of trailing Salt.

Signed-off-by: David Disseldorp <ddiss@suse.de>
Acked-by: Namjae Jeon <linkinjeon@kernel.org>
Cc: <stable@vger.kernel.org>
Signed-off-by: Steve French <stfrench@microsoft.com>
15 months agoMerge tag 'drm-misc-fixes-2023-04-13' of git://anongit.freedesktop.org/drm/drm-misc...
Daniel Vetter [Thu, 13 Apr 2023 18:47:58 +0000 (20:47 +0200)]
Merge tag 'drm-misc-fixes-2023-04-13' of git://anongit.freedesktop.org/drm/drm-misc into drm-fixes

Short summary of fixes pull:

 * armada: Fix double free
 * fb: Clear FB_ACTIVATE_KD_TEXT in ioctl
 * nouveau: Add missing callbacks
 * scheduler: Fix use-after-free error

Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
From: Thomas Zimmermann <tzimmermann@suse.de>
Link: https://patchwork.freedesktop.org/patch/msgid/20230413184233.GA8148@linux-uq9g
15 months agoMerge branch 'XDP-hints: change RX-hash kfunc bpf_xdp_metadata_rx_hash'
Alexei Starovoitov [Thu, 13 Apr 2023 18:05:49 +0000 (11:05 -0700)]
Merge branch 'XDP-hints: change RX-hash kfunc bpf_xdp_metadata_rx_hash'

Jesper Dangaard Brouer says:

====================

Current API for bpf_xdp_metadata_rx_hash() returns the raw RSS hash value,
but doesn't provide information on the RSS hash type (part of 6.3-rc).

This patchset proposal is to change the function call signature via adding
a pointer value argument for providing the RSS hash type.

Patchset also removes all bpf_printk's from xdp_hw_metadata program
that we expect driver developers to use. Instead counters are introduced
for relaying e.g. skip and fail info.
====================

Signed-off-by: Alexei Starovoitov <ast@kernel.org>
15 months agoselftests/bpf: Adjust bpf_xdp_metadata_rx_hash for new arg
Jesper Dangaard Brouer [Wed, 12 Apr 2023 19:49:00 +0000 (21:49 +0200)]
selftests/bpf: Adjust bpf_xdp_metadata_rx_hash for new arg

Update BPF selftests to use the new RSS type argument for kfunc
bpf_xdp_metadata_rx_hash.

Signed-off-by: Jesper Dangaard Brouer <brouer@redhat.com>
Acked-by: Toke Høiland-Jørgensen <toke@redhat.com>
Acked-by: Stanislav Fomichev <sdf@google.com>
Link: https://lore.kernel.org/r/168132894068.340624.8914711185697163690.stgit@firesoul
Signed-off-by: Alexei Starovoitov <ast@kernel.org>
15 months agomlx4: bpf_xdp_metadata_rx_hash add xdp rss hash type
Jesper Dangaard Brouer [Wed, 12 Apr 2023 19:48:55 +0000 (21:48 +0200)]
mlx4: bpf_xdp_metadata_rx_hash add xdp rss hash type

Update API for bpf_xdp_metadata_rx_hash() with arg for xdp rss hash type
via matching individual Completion Queue Entry (CQE) status bits.

Fixes: ab46182d0dcb ("net/mlx4_en: Support RX XDP metadata")
Signed-off-by: Jesper Dangaard Brouer <brouer@redhat.com>
Acked-by: Toke Høiland-Jørgensen <toke@redhat.com>
Acked-by: Stanislav Fomichev <sdf@google.com>
Link: https://lore.kernel.org/r/168132893562.340624.12779118462402031248.stgit@firesoul
Signed-off-by: Alexei Starovoitov <ast@kernel.org>
15 months agoveth: bpf_xdp_metadata_rx_hash add xdp rss hash type
Jesper Dangaard Brouer [Wed, 12 Apr 2023 19:48:50 +0000 (21:48 +0200)]
veth: bpf_xdp_metadata_rx_hash add xdp rss hash type

Update API for bpf_xdp_metadata_rx_hash() with arg for xdp rss hash type.

The veth driver currently only support XDP-hints based on SKB code path.
The SKB have lost information about the RSS hash type, by compressing
the information down to a single bitfield skb->l4_hash, that only knows
if this was a L4 hash value.

In preparation for veth, the xdp_rss_hash_type have an L4 indication
bit that allow us to return a meaningful L4 indication when working
with SKB based packets.

Fixes: 306531f0249f ("veth: Support RX XDP metadata")
Signed-off-by: Jesper Dangaard Brouer <brouer@redhat.com>
Acked-by: Toke Høiland-Jørgensen <toke@redhat.com>
Acked-by: Stanislav Fomichev <sdf@google.com>
Link: https://lore.kernel.org/r/168132893055.340624.16209448340644513469.stgit@firesoul
Signed-off-by: Alexei Starovoitov <ast@kernel.org>
15 months agomlx5: bpf_xdp_metadata_rx_hash add xdp rss hash type
Jesper Dangaard Brouer [Wed, 12 Apr 2023 19:48:45 +0000 (21:48 +0200)]
mlx5: bpf_xdp_metadata_rx_hash add xdp rss hash type

Update API for bpf_xdp_metadata_rx_hash() with arg for xdp rss hash type
via mapping table.

The mlx5 hardware can also identify and RSS hash IPSEC.  This indicate
hash includes SPI (Security Parameters Index) as part of IPSEC hash.

Extend xdp core enum xdp_rss_hash_type with IPSEC hash type.

Fixes: bc8d405b1ba9 ("net/mlx5e: Support RX XDP metadata")
Signed-off-by: Jesper Dangaard Brouer <brouer@redhat.com>
Acked-by: Toke Høiland-Jørgensen <toke@redhat.com>
Acked-by: Stanislav Fomichev <sdf@google.com>
Link: https://lore.kernel.org/r/168132892548.340624.11185734579430124869.stgit@firesoul
Signed-off-by: Alexei Starovoitov <ast@kernel.org>
15 months agoxdp: rss hash types representation
Jesper Dangaard Brouer [Wed, 12 Apr 2023 19:48:40 +0000 (21:48 +0200)]
xdp: rss hash types representation

The RSS hash type specifies what portion of packet data NIC hardware used
when calculating RSS hash value. The RSS types are focused on Internet
traffic protocols at OSI layers L3 and L4. L2 (e.g. ARP) often get hash
value zero and no RSS type. For L3 focused on IPv4 vs. IPv6, and L4
primarily TCP vs UDP, but some hardware supports SCTP.

Hardware RSS types are differently encoded for each hardware NIC. Most
hardware represent RSS hash type as a number. Determining L3 vs L4 often
requires a mapping table as there often isn't a pattern or sorting
according to ISO layer.

The patch introduce a XDP RSS hash type (enum xdp_rss_hash_type) that
contains both BITs for the L3/L4 types, and combinations to be used by
drivers for their mapping tables. The enum xdp_rss_type_bits get exposed
to BPF via BTF, and it is up to the BPF-programmer to match using these
defines.

This proposal change the kfunc API bpf_xdp_metadata_rx_hash() adding
a pointer value argument for provide the RSS hash type.
Change signature for all xmo_rx_hash calls in drivers to make it compile.

The RSS type implementations for each driver comes as separate patches.

Fixes: 3d76a4d3d4e5 ("bpf: XDP metadata RX kfuncs")
Signed-off-by: Jesper Dangaard Brouer <brouer@redhat.com>
Acked-by: Toke Høiland-Jørgensen <toke@redhat.com>
Acked-by: Stanislav Fomichev <sdf@google.com>
Link: https://lore.kernel.org/r/168132892042.340624.582563003880565460.stgit@firesoul
Signed-off-by: Alexei Starovoitov <ast@kernel.org>
15 months agoselftests/bpf: xdp_hw_metadata remove bpf_printk and add counters
Jesper Dangaard Brouer [Wed, 12 Apr 2023 19:48:35 +0000 (21:48 +0200)]
selftests/bpf: xdp_hw_metadata remove bpf_printk and add counters

The tool xdp_hw_metadata can be used by driver developers
implementing XDP-hints metadata kfuncs.

Remove all bpf_printk calls, as the tool already transfers all the
XDP-hints related information via metadata area to AF_XDP
userspace process.

Add counters for providing remaining information about failure and
skipped packet events.

Signed-off-by: Jesper Dangaard Brouer <brouer@redhat.com>
Acked-by: Stanislav Fomichev <sdf@google.com>
Link: https://lore.kernel.org/r/168132891533.340624.7313781245316405141.stgit@firesoul
Signed-off-by: Alexei Starovoitov <ast@kernel.org>
15 months agofbcon: set_con2fb_map needs to set con2fb_map!
Daniel Vetter [Wed, 12 Apr 2023 15:31:46 +0000 (17:31 +0200)]
fbcon: set_con2fb_map needs to set con2fb_map!

I got really badly confused in d443d9386472 ("fbcon: move more common
code into fb_open()") because we set the con2fb_map before the failure
points, which didn't look good.

But in trying to fix that I moved the assignment into the wrong path -
we need to do it for _all_ vc we take over, not just the first one
(which additionally requires the call to con2fb_acquire_newinfo).

I've figured this out because of a KASAN bug report, where the
fbcon_registered_fb and fbcon_display arrays went out of sync in
fbcon_mode_deleted() because the con2fb_map pointed at the old
fb_info, but the modes and everything was updated for the new one.

Signed-off-by: Daniel Vetter <daniel.vetter@intel.com>
Reviewed-by: Javier Martinez Canillas <javierm@redhat.com>
Acked-by: Helge Deller <deller@gmx.de>
Tested-by: Xingyuan Mo <hdthky0@gmail.com>
Fixes: d443d9386472 ("fbcon: move more common code into fb_open()")
Reported-by: Xingyuan Mo <hdthky0@gmail.com>
Cc: Thomas Zimmermann <tzimmermann@suse.de>
Cc: Sam Ravnborg <sam@ravnborg.org>
Cc: Xingyuan Mo <hdthky0@gmail.com>
Cc: Thomas Zimmermann <tzimmermann@suse.de>
Cc: Helge Deller <deller@gmx.de>
Cc: <stable@vger.kernel.org> # v5.19+
15 months agofbcon: Fix error paths in set_con2fb_map
Daniel Vetter [Wed, 12 Apr 2023 15:23:49 +0000 (17:23 +0200)]
fbcon: Fix error paths in set_con2fb_map

This is a regressoin introduced in b07db3958485 ("fbcon: Ditch error
handling for con2fb_release_oldinfo"). I failed to realize what the if
(!err) checks. The mentioned commit was dropping the
con2fb_release_oldinfo() return value but the if (!err) was also
checking whether the con2fb_acquire_newinfo() function call above
failed or not.

Fix this with an early return statement.

Note that there's still a difference compared to the orginal state of
the code, the below lines are now also skipped on error:

if (!search_fb_in_map(info_idx))
info_idx = newidx;

These are only needed when we've actually thrown out an old fb_info
from the console mappings, which only happens later on.

Also move the fbcon_add_cursor_work() call into the same if block,
it's all protected by console_lock so doesn't matter when we set up
the blinking cursor delayed work anyway. This further simplifies the
control flow and allows us to ditch the found local variable.

v2: Clarify commit message (Javier)

Signed-off-by: Daniel Vetter <daniel.vetter@intel.com>
Reviewed-by: Javier Martinez Canillas <javierm@redhat.com>
Acked-by: Helge Deller <deller@gmx.de>
Tested-by: Xingyuan Mo <hdthky0@gmail.com>
Fixes: b07db3958485 ("fbcon: Ditch error handling for con2fb_release_oldinfo")
Cc: Thomas Zimmermann <tzimmermann@suse.de>
Cc: Sam Ravnborg <sam@ravnborg.org>
Cc: Xingyuan Mo <hdthky0@gmail.com>
Cc: Thomas Zimmermann <tzimmermann@suse.de>
Cc: Helge Deller <deller@gmx.de>
Cc: <stable@vger.kernel.org> # v5.19+
15 months agoskbuff: Fix a race between coalescing and releasing SKBs
Liang Chen [Thu, 13 Apr 2023 09:03:53 +0000 (17:03 +0800)]
skbuff: Fix a race between coalescing and releasing SKBs

Commit 1effe8ca4e34 ("skbuff: fix coalescing for page_pool fragment
recycling") allowed coalescing to proceed with non page pool page and page
pool page when @from is cloned, i.e.

to->pp_recycle    --> false
from->pp_recycle  --> true
skb_cloned(from)  --> true

However, it actually requires skb_cloned(@from) to hold true until
coalescing finishes in this situation. If the other cloned SKB is
released while the merging is in process, from_shinfo->nr_frags will be
set to 0 toward the end of the function, causing the increment of frag
page _refcount to be unexpectedly skipped resulting in inconsistent
reference counts. Later when SKB(@to) is released, it frees the page
directly even though the page pool page is still in use, leading to
use-after-free or double-free errors. So it should be prohibited.

The double-free error message below prompted us to investigate:
BUG: Bad page state in process swapper/1  pfn:0e0d1
page:00000000c6548b28 refcount:-1 mapcount:0 mapping:0000000000000000
index:0x2 pfn:0xe0d1
flags: 0xfffffc0000000(node=0|zone=1|lastcpupid=0x1fffff)
raw: 000fffffc0000000 0000000000000000 ffffffff00000101 0000000000000000
raw: 0000000000000002 0000000000000000 ffffffffffffffff 0000000000000000
page dumped because: nonzero _refcount

CPU: 1 PID: 0 Comm: swapper/1 Tainted: G            E      6.2.0+
Call Trace:
 <IRQ>
dump_stack_lvl+0x32/0x50
bad_page+0x69/0xf0
free_pcp_prepare+0x260/0x2f0
free_unref_page+0x20/0x1c0
skb_release_data+0x10b/0x1a0
napi_consume_skb+0x56/0x150
net_rx_action+0xf0/0x350
? __napi_schedule+0x79/0x90
__do_softirq+0xc8/0x2b1
__irq_exit_rcu+0xb9/0xf0
common_interrupt+0x82/0xa0
</IRQ>
<TASK>
asm_common_interrupt+0x22/0x40
RIP: 0010:default_idle+0xb/0x20

Fixes: 53e0961da1c7 ("page_pool: add frag page recycling support in page pool")
Signed-off-by: Liang Chen <liangchen.linux@gmail.com>
Reviewed-by: Eric Dumazet <edumazet@google.com>
Link: https://lore.kernel.org/r/20230413090353.14448-1-liangchen.linux@gmail.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
15 months agonet: macb: fix a memory corruption in extended buffer descriptor mode
Roman Gushchin [Wed, 12 Apr 2023 23:21:44 +0000 (16:21 -0700)]
net: macb: fix a memory corruption in extended buffer descriptor mode

For quite some time we were chasing a bug which looked like a sudden
permanent failure of networking and mmc on some of our devices.
The bug was very sensitive to any software changes and even more to
any kernel debug options.

Finally we got a setup where the problem was reproducible with
CONFIG_DMA_API_DEBUG=y and it revealed the issue with the rx dma:

[   16.992082] ------------[ cut here ]------------
[   16.996779] DMA-API: macb ff0b0000.ethernet: device driver tries to free DMA memory it has not allocated [device address=0x0000000875e3e244] [size=1536 bytes]
[   17.011049] WARNING: CPU: 0 PID: 85 at kernel/dma/debug.c:1011 check_unmap+0x6a0/0x900
[   17.018977] Modules linked in: xxxxx
[   17.038823] CPU: 0 PID: 85 Comm: irq/55-8000f000 Not tainted 5.4.0 #28
[   17.045345] Hardware name: xxxxx
[   17.049528] pstate: 60000005 (nZCv daif -PAN -UAO)
[   17.054322] pc : check_unmap+0x6a0/0x900
[   17.058243] lr : check_unmap+0x6a0/0x900
[   17.062163] sp : ffffffc010003c40
[   17.065470] x29: ffffffc010003c40 x28: 000000004000c03c
[   17.070783] x27: ffffffc010da7048 x26: ffffff8878e38800
[   17.076095] x25: ffffff8879d22810 x24: ffffffc010003cc8
[   17.081407] x23: 0000000000000000 x22: ffffffc010a08750
[   17.086719] x21: ffffff8878e3c7c0 x20: ffffffc010acb000
[   17.092032] x19: 0000000875e3e244 x18: 0000000000000010
[   17.097343] x17: 0000000000000000 x16: 0000000000000000
[   17.102647] x15: ffffff8879e4a988 x14: 0720072007200720
[   17.107959] x13: 0720072007200720 x12: 0720072007200720
[   17.113261] x11: 0720072007200720 x10: 0720072007200720
[   17.118565] x9 : 0720072007200720 x8 : 000000000000022d
[   17.123869] x7 : 0000000000000015 x6 : 0000000000000098
[   17.129173] x5 : 0000000000000000 x4 : 0000000000000000
[   17.134475] x3 : 00000000ffffffff x2 : ffffffc010a1d370
[   17.139778] x1 : b420c9d75d27bb00 x0 : 0000000000000000
[   17.145082] Call trace:
[   17.147524]  check_unmap+0x6a0/0x900
[   17.151091]  debug_dma_unmap_page+0x88/0x90
[   17.155266]  gem_rx+0x114/0x2f0
[   17.158396]  macb_poll+0x58/0x100
[   17.161705]  net_rx_action+0x118/0x400
[   17.165445]  __do_softirq+0x138/0x36c
[   17.169100]  irq_exit+0x98/0xc0
[   17.172234]  __handle_domain_irq+0x64/0xc0
[   17.176320]  gic_handle_irq+0x5c/0xc0
[   17.179974]  el1_irq+0xb8/0x140
[   17.183109]  xiic_process+0x5c/0xe30
[   17.186677]  irq_thread_fn+0x28/0x90
[   17.190244]  irq_thread+0x208/0x2a0
[   17.193724]  kthread+0x130/0x140
[   17.196945]  ret_from_fork+0x10/0x20
[   17.200510] ---[ end trace 7240980785f81d6f ]---

[  237.021490] ------------[ cut here ]------------
[  237.026129] DMA-API: exceeded 7 overlapping mappings of cacheline 0x0000000021d79e7b
[  237.033886] WARNING: CPU: 0 PID: 0 at kernel/dma/debug.c:499 add_dma_entry+0x214/0x240
[  237.041802] Modules linked in: xxxxx
[  237.061637] CPU: 0 PID: 0 Comm: swapper/0 Tainted: G        W         5.4.0 #28
[  237.068941] Hardware name: xxxxx
[  237.073116] pstate: 80000085 (Nzcv daIf -PAN -UAO)
[  237.077900] pc : add_dma_entry+0x214/0x240
[  237.081986] lr : add_dma_entry+0x214/0x240
[  237.086072] sp : ffffffc010003c30
[  237.089379] x29: ffffffc010003c30 x28: ffffff8878a0be00
[  237.094683] x27: 0000000000000180 x26: ffffff8878e387c0
[  237.099987] x25: 0000000000000002 x24: 0000000000000000
[  237.105290] x23: 000000000000003b x22: ffffffc010a0fa00
[  237.110594] x21: 0000000021d79e7b x20: ffffffc010abe600
[  237.115897] x19: 00000000ffffffef x18: 0000000000000010
[  237.121201] x17: 0000000000000000 x16: 0000000000000000
[  237.126504] x15: ffffffc010a0fdc8 x14: 0720072007200720
[  237.131807] x13: 0720072007200720 x12: 0720072007200720
[  237.137111] x11: 0720072007200720 x10: 0720072007200720
[  237.142415] x9 : 0720072007200720 x8 : 0000000000000259
[  237.147718] x7 : 0000000000000001 x6 : 0000000000000000
[  237.153022] x5 : ffffffc010003a20 x4 : 0000000000000001
[  237.158325] x3 : 0000000000000006 x2 : 0000000000000007
[  237.163628] x1 : 8ac721b3a7dc1c00 x0 : 0000000000000000
[  237.168932] Call trace:
[  237.171373]  add_dma_entry+0x214/0x240
[  237.175115]  debug_dma_map_page+0xf8/0x120
[  237.179203]  gem_rx_refill+0x190/0x280
[  237.182942]  gem_rx+0x224/0x2f0
[  237.186075]  macb_poll+0x58/0x100
[  237.189384]  net_rx_action+0x118/0x400
[  237.193125]  __do_softirq+0x138/0x36c
[  237.196780]  irq_exit+0x98/0xc0
[  237.199914]  __handle_domain_irq+0x64/0xc0
[  237.204000]  gic_handle_irq+0x5c/0xc0
[  237.207654]  el1_irq+0xb8/0x140
[  237.210789]  arch_cpu_idle+0x40/0x200
[  237.214444]  default_idle_call+0x18/0x30
[  237.218359]  do_idle+0x200/0x280
[  237.221578]  cpu_startup_entry+0x20/0x30
[  237.225493]  rest_init+0xe4/0xf0
[  237.228713]  arch_call_rest_init+0xc/0x14
[  237.232714]  start_kernel+0x47c/0x4a8
[  237.236367] ---[ end trace 7240980785f81d70 ]---

Lars was fast to find an explanation: according to the datasheet
bit 2 of the rx buffer descriptor entry has a different meaning in the
extended mode:
  Address [2] of beginning of buffer, or
  in extended buffer descriptor mode (DMA configuration register [28] = 1),
  indicates a valid timestamp in the buffer descriptor entry.

The macb driver didn't mask this bit while getting an address and it
eventually caused a memory corruption and a dma failure.

The problem is resolved by explicitly clearing the problematic bit
if hw timestamping is used.

Fixes: 7b4296148066 ("net: macb: Add support for PTP timestamps in DMA descriptors")
Signed-off-by: Roman Gushchin <roman.gushchin@linux.dev>
Co-developed-by: Lars-Peter Clausen <lars@metafoo.de>
Signed-off-by: Lars-Peter Clausen <lars@metafoo.de>
Acked-by: Nicolas Ferre <nicolas.ferre@microchip.com>
Reviewed-by: Jacob Keller <jacob.e.keller@intel.com>
Link: https://lore.kernel.org/r/20230412232144.770336-1-roman.gushchin@linux.dev
Signed-off-by: Jakub Kicinski <kuba@kernel.org>