Currently the dma debugging code can end up indirectly calling printk
under the radix_lock. This happens when a radix tree node allocation
fails.
This is a problem because the printk code, when used together with
netconsole, can end up inside the dma debugging code while trying to
transmit a message over netcons.
This creates the possibility of either a circular deadlock on the same
CPU, with that CPU trying to grab the radix_lock twice, or an ABBA
deadlock between different CPUs, where one CPU grabs the console lock
first and then waits for the radix_lock, while the other CPU is holding
the radix_lock and is waiting for the console lock.
The trace captured by lockdep is of the ABBA variant.
-> #2 (&dma_entry_hash[i].lock){-.-.}-{2:2}:
_raw_spin_lock_irqsave+0x5a/0x90
debug_dma_map_page+0x79/0x180
dma_map_page_attrs+0x1d2/0x2f0
bnxt_start_xmit+0x8c6/0x1540
netpoll_start_xmit+0x13f/0x180
netpoll_send_skb+0x20d/0x320
netpoll_send_udp+0x453/0x4a0
write_ext_msg+0x1b9/0x460
console_flush_all+0x2ff/0x5a0
console_unlock+0x55/0x180
vprintk_emit+0x2e3/0x3c0
devkmsg_emit+0x5a/0x80
devkmsg_write+0xfd/0x180
do_iter_readv_writev+0x164/0x1b0
vfs_writev+0xf9/0x2b0
do_writev+0x6d/0x110
do_syscall_64+0x80/0x150
entry_SYSCALL_64_after_hwframe+0x4b/0x53
-> #0 (console_owner){-.-.}-{0:0}:
__lock_acquire+0x15d1/0x31a0
lock_acquire+0xe8/0x290
console_flush_all+0x2ea/0x5a0
console_unlock+0x55/0x180
vprintk_emit+0x2e3/0x3c0
_printk+0x59/0x80
warn_alloc+0x122/0x1b0
__alloc_pages_slowpath+0x1101/0x1120
__alloc_pages+0x1eb/0x2c0
alloc_slab_page+0x5f/0x150
new_slab+0x2dc/0x4e0
___slab_alloc+0xdcb/0x1390
kmem_cache_alloc+0x23d/0x360
radix_tree_node_alloc+0x3c/0xf0
radix_tree_insert+0xf5/0x230
add_dma_entry+0xe9/0x360
dma_map_page_attrs+0x1d2/0x2f0
__bnxt_alloc_rx_frag+0x147/0x180
bnxt_alloc_rx_data+0x79/0x160
bnxt_rx_skb+0x29/0xc0
bnxt_rx_pkt+0xe22/0x1570
__bnxt_poll_work+0x101/0x390
bnxt_poll+0x7e/0x320
__napi_poll+0x29/0x160
net_rx_action+0x1e0/0x3e0
handle_softirqs+0x190/0x510
run_ksoftirqd+0x4e/0x90
smpboot_thread_fn+0x1a8/0x270
kthread+0x102/0x120
ret_from_fork+0x2f/0x40
ret_from_fork_asm+0x11/0x20
This bug is more likely than it seems, because when one CPU has run out
of memory, chances are the other has too.
The good news is, this bug is hidden behind the CONFIG_DMA_API_DEBUG, so
not many users are likely to trigger it.
Signed-off-by: Rik van Riel <riel@surriel.com>
Reported-by: Konstantin Ovsepian <ovs@meta.com>
Signed-off-by: Christoph Hellwig <hch@lst.de>
Update the kernel-doc comments to catch up with the code changes and
fix the kernel-doc warnings:
debug.c:83: warning: Excess struct member 'stacktrace' description in 'dma_debug_entry'
debug.c:83: warning: Function parameter or struct member 'stack_len' not described in 'dma_debug_entry'
debug.c:83: warning: Function parameter or struct member 'stack_entries' not described in 'dma_debug_entry'
Fixes: 746017ed8d ("dma/debug: Simplify stracktrace retrieval")
Signed-off-by: Randy Dunlap <rdunlap@infradead.org>
Cc: Christoph Hellwig <hch@lst.de>
Cc: Marek Szyprowski <m.szyprowski@samsung.com>
Cc: Robin Murphy <robin.murphy@arm.com>
Cc: iommu@lists.linux.dev
Signed-off-by: Christoph Hellwig <hch@lst.de>
The driver core now can handle a const struct bus_type pointer, and the
dma_debug_add_bus() call just passes on the pointer give to it to the
driver core, so make this pointer const as well to allow everyone to use
read-only struct bus_type pointers going forward.
Cc: Christoph Hellwig <hch@lst.de>
Cc: Marek Szyprowski <m.szyprowski@samsung.com>
Cc: Robin Murphy <robin.murphy@arm.com>
Cc: <iommu@lists.linux.dev>
Reviewed-by: Robin Murphy <robin.murphy@arm.com>
Link: https://lore.kernel.org/r/2023121941-dejected-nugget-681e@gregkh
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
On 32-bit without LPAE:
kernel/dma/debug.c: In function ‘debug_dma_dump_mappings’:
kernel/dma/debug.c:537:7: warning: format ‘%llx’ expects argument of type ‘long long unsigned int’, but argument 9 has type ‘phys_addr_t’ {aka ‘unsigned int’} [-Wformat=]
kernel/dma/debug.c: In function ‘dump_show’:
kernel/dma/debug.c:568:59: warning: format ‘%llx’ expects argument of type ‘long long unsigned int’, but argument 11 has type ‘phys_addr_t’ {aka ‘unsigned int’} [-Wformat=]
Fixes: bd89d69a52 ("dma-debug: add cacheline to user/kernel space dump messages")
Reported-by: kernel test robot <lkp@intel.com>
Link: https://lore.kernel.org/r/202303160548.ReyuTsGD-lkp@intel.com
Reported-by: noreply@ellerman.id.au
Signed-off-by: Geert Uytterhoeven <geert+renesas@glider.be>
Signed-off-by: Christoph Hellwig <hch@lst.de>
Having the cacheline also printed on the debug_dma_dump_mappings() and
dump_show() is useful for debugging. Furthermore, this also standardizes
the messages shown on both dump functions.
Signed-off-by: Desnes Nunes <desnesn@redhat.com>
Signed-off-by: Christoph Hellwig <hch@lst.de>
Small update on dma_debug_entry's struct commentary and also standardize
the usage of 'dma_addr' variable name from debug_dma_map_page() on
debug_dma_unmap_page(), and similarly on debug_dma_free_coherent()
Signed-off-by: Desnes Nunes <desnesn@redhat.com>
Signed-off-by: Christoph Hellwig <hch@lst.de>
When bucket_find_contains() tries to find the original entry for a
partial sync, it manages to constrain its search in a way that is both
too restrictive and not restrictive enough. A driver which only uses
single mappings rather than scatterlists might not set max_seg_size, but
could still technically perform a partial sync at an offset of more than
64KB into a sufficiently large mapping, so we could stop searching too
early before reaching a legitimate entry. Conversely, if no valid entry
is present and max_range is large enough, we can pointlessly search
buckets that we've already searched, or that represent an impossible
wrapping around the bottom of the address space. At worst, the
(legitimate) case of max_seg_size == UINT_MAX can make the loop
infinite.
Replace the fragile and frankly hard-to-follow "range" logic with a
simple counted loop for the number of possible hash buckets below the
given address.
Reported-by: Yunfei Wang <yf.wang@mediatek.com>
Signed-off-by: Robin Murphy <robin.murphy@arm.com>
Signed-off-by: Christoph Hellwig <hch@lst.de>
Limit the error msg to avoid flooding the console. If you have a lot of
threads hitting this at once, they could have already gotten passed the
dma_debug_disabled() check before they get to the point of allocation
failure, resulting in quite a lot of this error message spamming the
log. Use pr_err_once() to limit that.
Signed-off-by: Rob Clark <robdclark@chromium.org>
Signed-off-by: Christoph Hellwig <hch@lst.de>
We observed the error "cacheline tracking ENOMEM, dma-debug disabled"
during a light system load (copying some files). The reason for this error
is that the dma_active_cacheline radix tree uses GFP_NOWAIT allocation -
so it can't access the emergency memory reserves and it fails as soon as
anybody reaches the watermark.
This patch changes GFP_NOWAIT to GFP_ATOMIC, so that it can access the
emergency memory reserves.
Signed-off-by: Mikulas Patocka <mpatocka@redhat.com>
Signed-off-by: Christoph Hellwig <hch@lst.de>
When valid kernel command line parameters
dma_debug=off dma_debug_entries=100
are used, they are reported as Unknown parameters and added to init's
environment strings, polluting it.
Unknown kernel command line parameters "BOOT_IMAGE=/boot/bzImage-517rc5
dma_debug=off dma_debug_entries=100", will be passed to user space.
and
Run /sbin/init as init process
with arguments:
/sbin/init
with environment:
HOME=/
TERM=linux
BOOT_IMAGE=/boot/bzImage-517rc5
dma_debug=off
dma_debug_entries=100
Return 1 from these __setup handlers to indicate that the command line
option has been handled.
Fixes: 59d3daafa1 ("dma-debug: add kernel command line parameters")
Signed-off-by: Randy Dunlap <rdunlap@infradead.org>
Reported-by: Igor Zhbanov <i.zhbanov@omprussia.ru>
Link: lore.kernel.org/r/64644a2f-4a20-bab3-1e15-3b2cdd0defe3@omprussia.ru
Cc: Joerg Roedel <joro@8bytes.org>
Cc: Christoph Hellwig <hch@lst.de>
Cc: Marek Szyprowski <m.szyprowski@samsung.com>
Cc: iommu@lists.linux-foundation.org
Cc: Robin Murphy <robin.murphy@arm.com>
Signed-off-by: Christoph Hellwig <hch@lst.de>
Mapping something twice should be possible as long as,
DMA_ATTR_SKIP_CPU_SYNC is passed to the strictly speaking second relevant
mapping operation (that attempts to map the same thing). So, don't issue a
warning if the specified condition is met in add_dma_entry().
Signed-off-by: Hamza Mahfooz <someguy@effective-light.com>
Signed-off-by: Christoph Hellwig <hch@lst.de>
The following warning occurred sporadically on s390:
DMA-API: nvme 0006:00:00.0: device driver maps memory from kernel text or rodata [addr=0000000048cc5e2f] [len=131072]
WARNING: CPU: 4 PID: 825 at kernel/dma/debug.c:1083 check_for_illegal_area+0xa8/0x138
It is a false-positive warning, due to broken logic in debug_dma_map_sg().
check_for_illegal_area() checks for overlay of sg elements with kernel text
or rodata. It is called with sg_dma_len(s) instead of s->length as
parameter. After the call to ->map_sg(), sg_dma_len() will contain the
length of possibly combined sg elements in the DMA address space, and not
the individual sg element length, which would be s->length.
The check will then use the physical start address of an sg element, and
add the DMA length for the overlap check, which could result in the false
warning, because the DMA length can be larger than the actual single sg
element length.
In addition, the call to check_for_illegal_area() happens in the iteration
over mapped_ents, which will not include all individual sg elements if
any of them were combined in ->map_sg().
Fix this by using s->length instead of sg_dma_len(s). Also put the call to
check_for_illegal_area() in a separate loop, iterating over all the
individual sg elements ("nents" instead of "mapped_ents").
While at it, as suggested by Robin Murphy, also move check_for_stack()
inside the new loop, as it is similarly concerned with validating the
individual sg elements.
Link: https://lore.kernel.org/lkml/20210705185252.4074653-1-gerald.schaefer@linux.ibm.com
Fixes: 884d05970b ("dma-debug: use sg_dma_len accessor")
Signed-off-by: Gerald Schaefer <gerald.schaefer@linux.ibm.com>
Signed-off-by: Christoph Hellwig <hch@lst.de>
For some drivers, that use the DMA API. This error message can be reached
several millions of times per second, causing spam to the kernel's printk
buffer and bringing the CPU usage up to 100% (so, it should be rate
limited). However, since there is at least one driver that is in the
mainline and suffers from the error condition, it is more useful to
err_printk() here instead of just rate limiting the error message (in hopes
that it will make it easier for other drivers that suffer from this issue
to be spotted).
Link: https://lkml.kernel.org/r/fd67fbac-64bf-f0ea-01e1-5938ccfab9d0@arm.com
Reported-by: Jeremy Linton <jeremy.linton@arm.com>
Signed-off-by: Hamza Mahfooz <someguy@effective-light.com>
Signed-off-by: Christoph Hellwig <hch@lst.de>
Due to link order, dma_debug_init is called before debugfs has a chance
to initialize (via debugfs_init which also happens in the core initcall
stage), so the directories for dma-debug are never created.
Decouple dma_debug_fs_init from dma_debug_init and defer its init until
core_initcall_sync (after debugfs has been initialized) while letting
dma-debug initialization occur as soon as possible to catch any early
mappings, as suggested in [1].
[1] https://lore.kernel.org/linux-iommu/YIgGa6yF%2Fadg8OSN@kroah.com/
Fixes: 15b28bbcd5 ("dma-debug: move initialization to common code")
Signed-off-by: Anthony Iliopoulos <ailiop@suse.com>
Signed-off-by: Christoph Hellwig <hch@lst.de>
There is already a memory_intersects() helper in sections.h,
use memory_intersects() directly instead of private overlap().
Signed-off-by: Kefeng Wang <wangkefeng.wang@huawei.com>
Signed-off-by: Christoph Hellwig <hch@lst.de>
Since, overlapping mappings are not supported by the DMA API we should
report an error if active_cacheline_insert returns -EEXIST.
Suggested-by: Dan Williams <dan.j.williams@intel.com>
Signed-off-by: Hamza Mahfooz <someguy@effective-light.com>
Signed-off-by: Christoph Hellwig <hch@lst.de>
Most of dma-debug.h is not required by anything outside of kernel/dma.
Move the four declarations needed by dma-mappin.h or dma-ops providers
into dma-mapping.h and dma-map-ops.h, and move the remainder of the
file to kernel/dma/debug.h.
Signed-off-by: Christoph Hellwig <hch@lst.de>
Now that the main dma mapping entry points are out of line most of the
symbols in dma-debug.c can only be called from built-in code. Remove
the unused exports.
Signed-off-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: Robin Murphy <robin.murphy@arm.com>
This remoes the code from the COW path to call debug_dma_assert_idle(),
which was added many years ago.
Google shows that it hasn't caught anything in the 6+ years we've had it
apart from a false positive, and Hugh just noticed how it had a very
unfortunate spinlock serialization in the COW path.
He fixed that issue the previous commit (a85ffd59bd: "dma-debug: fix
debug_dma_assert_idle(), use rcu_read_lock()"), but let's see if anybody
even notices when we remove this function entirely.
NOTE! We keep the dma tracking infrastructure that was added by the
commit that introduced it. Partly to make it easier to resurrect this
debug code if we ever deside to, and partly because that tracking by pfn
and offset looks quite reasonable.
The problem with this debug code was simply that it was expensive and
didn't seem worth it, not that it was wrong per se.
Acked-by: Dan Williams <dan.j.williams@intel.com>
Acked-by: Christoph Hellwig <hch@lst.de>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Since commit 2a9127fcf2 ("mm: rewrite wait_on_page_bit_common()
logic") improved unlock_page(), it has become more noticeable how
cow_user_page() in a kernel with CONFIG_DMA_API_DEBUG=y can create and
suffer from heavy contention on DMA debug's radix_lock in
debug_dma_assert_idle().
It is only doing a lookup: use rcu_read_lock() and rcu_read_unlock()
instead; though that does require the static ents[] to be moved
onstack...
...but, hold on, isn't that radix_tree_gang_lookup() and loop doing
quite the wrong thing: searching CACHELINES_PER_PAGE entries for an
exact match with the first cacheline of the page in question?
radix_tree_gang_lookup() is the right tool for the job, but we need
nothing more than to check the first entry it can find, reporting if
that falls anywhere within the page.
(Is RCU safe here? As safe as using the spinlock was. The entries are
never freed, so don't need to be freed by RCU. They may be reused, and
there is a faint chance of a race, with an offending entry reused while
printing its error info; but the spinlock did not prevent that either,
and I agree that it's not worth worrying about. ]
[ Side noe: this patch is a clear improvement to the status quo, but the
next patch will be removing this debug function entirely.
But just in case we decide we want to resurrect the debugging code
some day, I'm first applying this improvement patch so that it doesn't
get lost - Linus ]
Fixes: 3b7a6418c7 ("dma debug: account for cachelines and read-only mappings in overlap tracking")
Signed-off-by: Hugh Dickins <hughd@google.com>
Acked-by: Dan Williams <dan.j.williams@intel.com>
Acked-by: Christoph Hellwig <hch@lst.de>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
while to come. Changes include:
- Some new Chinese translations
- Progress on the battle against double words words and non-HTTPS URLs
- Some block-mq documentation
- More RST conversions from Mauro. At this point, that task is
essentially complete, so we shouldn't see this kind of churn again for a
while. Unless we decide to switch to asciidoc or something...:)
- Lots of typo fixes, warning fixes, and more.
-----BEGIN PGP SIGNATURE-----
iQFDBAABCAAtFiEEIw+MvkEiF49krdp9F0NaE2wMflgFAl8oVkwPHGNvcmJldEBs
d24ubmV0AAoJEBdDWhNsDH5YoW8H/jJ/xnXFn7tkgVPQAlL3k5HCnK7A5nDP9RVR
cg1pTx1cEFdjzxPlJyExU6/v+AImOvtweHXC+JDK7YcJ6XFUNYXJI3LxL5KwUXbY
BL/xRFszDSXH2C7SJF5GECcFYp01e/FWSLN3yWAh+g+XwsKiTJ8q9+CoIDkHfPGO
7oQsHKFu6s36Af0LfSgxk4sVB7EJbo8e4psuPsP5SUrl+oXRO43Put0rXkR4yJoH
9oOaB51Do5fZp8I4JVAqGXvpXoExyLMO4yw0mASm6YSZ3KyjR8Fae+HD9Cq4ZuwY
0uzb9K+9NEhqbfwtyBsi99S64/6Zo/MonwKwevZuhtsDTK4l4iU=
=JQLZ
-----END PGP SIGNATURE-----
Merge tag 'docs-5.9' of git://git.lwn.net/linux
Pull documentation updates from Jonathan Corbet:
"It's been a busy cycle for documentation - hopefully the busiest for a
while to come. Changes include:
- Some new Chinese translations
- Progress on the battle against double words words and non-HTTPS
URLs
- Some block-mq documentation
- More RST conversions from Mauro. At this point, that task is
essentially complete, so we shouldn't see this kind of churn again
for a while. Unless we decide to switch to asciidoc or
something...:)
- Lots of typo fixes, warning fixes, and more"
* tag 'docs-5.9' of git://git.lwn.net/linux: (195 commits)
scripts/kernel-doc: optionally treat warnings as errors
docs: ia64: correct typo
mailmap: add entry for <alobakin@marvell.com>
doc/zh_CN: add cpu-load Chinese version
Documentation/admin-guide: tainted-kernels: fix spelling mistake
MAINTAINERS: adjust kprobes.rst entry to new location
devices.txt: document rfkill allocation
PCI: correct flag name
docs: filesystems: vfs: correct flag name
docs: filesystems: vfs: correct sync_mode flag names
docs: path-lookup: markup fixes for emphasis
docs: path-lookup: more markup fixes
docs: path-lookup: fix HTML entity mojibake
CREDITS: Replace HTTP links with HTTPS ones
docs: process: Add an example for creating a fixes tag
doc/zh_CN: add Chinese translation prefer section
doc/zh_CN: add clearing-warn-once Chinese version
doc/zh_CN: add admin-guide index
doc:it_IT: process: coding-style.rst: Correct __maybe_unused compiler label
futex: MAINTAINERS: Re-add selftests directory
...
- make support for dma_ops optional
- move more code out of line
- add generic support for a dma_ops bypass mode
- misc cleanups
-----BEGIN PGP SIGNATURE-----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=90UK
-----END PGP SIGNATURE-----
Merge tag 'dma-mapping-5.9' of git://git.infradead.org/users/hch/dma-mapping
Pull dma-mapping updates from Christoph Hellwig:
- make support for dma_ops optional
- move more code out of line
- add generic support for a dma_ops bypass mode
- misc cleanups
* tag 'dma-mapping-5.9' of git://git.infradead.org/users/hch/dma-mapping:
dma-contiguous: cleanup dma_alloc_contiguous
dma-debug: use named initializers for dir2name
powerpc: use the generic dma_ops_bypass mode
dma-mapping: add a dma_ops_bypass flag to struct device
dma-mapping: make support for dma ops optional
dma-mapping: inline the fast path dma-direct calls
dma-mapping: move the remaining DMA API calls out of line
Make dir2name a little more readable and maintainable by using
named initializers.
Signed-off-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: Robin Murphy <robin.murphy@arm.com>
Using uninitialized_var() is dangerous as it papers over real bugs[1]
(or can in the future), and suppresses unrelated compiler warnings
(e.g. "unused variable"). If the compiler thinks it is uninitialized,
either simply initialize the variable or make compiler changes.
In preparation for removing[2] the[3] macro[4], remove all remaining
needless uses with the following script:
git grep '\buninitialized_var\b' | cut -d: -f1 | sort -u | \
xargs perl -pi -e \
's/\buninitialized_var\(([^\)]+)\)/\1/g;
s:\s*/\* (GCC be quiet|to make compiler happy) \*/$::g;'
drivers/video/fbdev/riva/riva_hw.c was manually tweaked to avoid
pathological white-space.
No outstanding warnings were found building allmodconfig with GCC 9.3.0
for x86_64, i386, arm64, arm, powerpc, powerpc64le, s390x, mips, sparc64,
alpha, and m68k.
[1] https://lore.kernel.org/lkml/20200603174714.192027-1-glider@google.com/
[2] https://lore.kernel.org/lkml/CA+55aFw+Vbj0i=1TGqCR5vQkCzWJ0QxK6CernOU6eedsudAixw@mail.gmail.com/
[3] https://lore.kernel.org/lkml/CA+55aFwgbgqhbp1fkxvRKEpzyR5J8n1vKT1VZdz9knmPuXhOeg@mail.gmail.com/
[4] https://lore.kernel.org/lkml/CA+55aFz2500WfbKXAx8s67wrm9=yVJu65TpLgN_ybYNv0VEOKA@mail.gmail.com/
Reviewed-by: Leon Romanovsky <leonro@mellanox.com> # drivers/infiniband and mlx4/mlx5
Acked-by: Jason Gunthorpe <jgg@mellanox.com> # IB
Acked-by: Kalle Valo <kvalo@codeaurora.org> # wireless drivers
Reviewed-by: Chao Yu <yuchao0@huawei.com> # erofs
Signed-off-by: Kees Cook <keescook@chromium.org>
Fix the following sparse warning:
kernel/dma/debug.c:659:6: warning: symbol '__dma_entry_alloc_check_leak'
was not declared. Should it be static?
Reported-by: Hulk Robot <hulkci@huawei.com>
Signed-off-by: Jason Yan <yanaijie@huawei.com>
Signed-off-by: Christoph Hellwig <hch@lst.de>
- improve dma-debug scalability (Eric Dumazet)
- tiny dma-debug cleanup (Dan Carpenter)
- check for vmap memory in dma_map_single (Kees Cook)
- check for dma_addr_t overflows in dma-direct when using
DMA offsets (Nicolas Saenz Julienne)
- switch the x86 sta2x11 SOC to use more generic DMA code
(Nicolas Saenz Julienne)
- fix arm-nommu dma-ranges handling (Vladimir Murzin)
- use __initdata in CMA (Shyam Saini)
- replace the bus dma mask with a limit (Nicolas Saenz Julienne)
- merge the remapping helpers into the main dma-direct flow (me)
- switch xtensa to the generic dma remap handling (me)
- various cleanups around dma_capable (me)
- remove unused dev arguments to various dma-noncoherent helpers (me)
-----BEGIN PGP SIGNATURE-----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=e5wO
-----END PGP SIGNATURE-----
Merge branch 'master' of git://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux; tag 'dma-mapping-5.5' of git://git.infradead.org/users/hch/dma-mapping
Pull dma-mapping updates from Christoph Hellwig:
- improve dma-debug scalability (Eric Dumazet)
- tiny dma-debug cleanup (Dan Carpenter)
- check for vmap memory in dma_map_single (Kees Cook)
- check for dma_addr_t overflows in dma-direct when using DMA offsets
(Nicolas Saenz Julienne)
- switch the x86 sta2x11 SOC to use more generic DMA code (Nicolas
Saenz Julienne)
- fix arm-nommu dma-ranges handling (Vladimir Murzin)
- use __initdata in CMA (Shyam Saini)
- replace the bus dma mask with a limit (Nicolas Saenz Julienne)
- merge the remapping helpers into the main dma-direct flow (me)
- switch xtensa to the generic dma remap handling (me)
- various cleanups around dma_capable (me)
- remove unused dev arguments to various dma-noncoherent helpers (me)
* 'master' of git://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux:
* tag 'dma-mapping-5.5' of git://git.infradead.org/users/hch/dma-mapping: (22 commits)
dma-mapping: treat dev->bus_dma_mask as a DMA limit
dma-direct: exclude dma_direct_map_resource from the min_low_pfn check
dma-direct: don't check swiotlb=force in dma_direct_map_resource
dma-debug: clean up put_hash_bucket()
powerpc: remove support for NULL dev in __phys_to_dma / __dma_to_phys
dma-direct: avoid a forward declaration for phys_to_dma
dma-direct: unify the dma_capable definitions
dma-mapping: drop the dev argument to arch_sync_dma_for_*
x86/PCI: sta2x11: use default DMA address translation
dma-direct: check for overflows on 32 bit DMA addresses
dma-debug: increase HASH_SIZE
dma-debug: reorder struct dma_debug_entry fields
xtensa: use the generic uncached segment support
dma-mapping: merge the generic remapping helpers into dma-direct
dma-direct: provide mmap and get_sgtable method overrides
dma-direct: remove the dma_handle argument to __dma_direct_alloc_pages
dma-direct: remove __dma_direct_free_pages
usb: core: Remove redundant vmap checks
kernel: dma-contiguous: mark CMA parameters __initdata/__initconst
dma-debug: add a schedule point in debug_dma_dump_mappings()
...
The put_hash_bucket() is a bit cleaner if it takes an unsigned long
directly instead of a pointer to unsigned long.
Signed-off-by: Dan Carpenter <dan.carpenter@oracle.com>
Signed-off-by: Christoph Hellwig <hch@lst.de>
With modern NIC, it is not unusual having about ~256,000 active dma
mappings and a hash size of 1024 buckets is too small.
Forcing full cache line per bucket does not seem useful, especially now
that we have contention on free_entries_lock for allocations and freeing
of entries. Better use the space to fit more buckets.
Signed-off-by: Eric Dumazet <edumazet@google.com>
Signed-off-by: Christoph Hellwig <hch@lst.de>
Move all fields used during exact match lookups to the first cache line.
This makes debug_dma_mapping_error() and friends about 50% faster.
Since it removes two 32bit holes, force a cacheline alignment on struct
dma_debug_entry.
Signed-off-by: Eric Dumazet <edumazet@google.com>
Signed-off-by: Christoph Hellwig <hch@lst.de>
debug_dma_dump_mappings() can take a lot of cpu cycles :
lpk43:/# time wc -l /sys/kernel/debug/dma-api/dump
163435 /sys/kernel/debug/dma-api/dump
real 0m0.463s
user 0m0.003s
sys 0m0.459s
Let's add a cond_resched() to avoid holding cpu for too long.
Signed-off-by: Eric Dumazet <edumazet@google.com>
Cc: Corentin Labbe <clabbe@baylibre.com>
Cc: Christoph Hellwig <hch@lst.de>
Cc: Marek Szyprowski <m.szyprowski@samsung.com>
Signed-off-by: Christoph Hellwig <hch@lst.de>
As said in commit f2c2cbcc35 ("powerpc: Use pr_warn instead of
pr_warning"), removing pr_warning so all logging messages use a
consistent <prefix>_warn style. Let's do it.
Link: http://lkml.kernel.org/r/20191018031850.48498-25-wangkefeng.wang@huawei.com
To: linux-kernel@vger.kernel.org
Cc: Christoph Hellwig <hch@lst.de>
Cc: Marek Szyprowski <m.szyprowski@samsung.com>
Cc: Robin Murphy <robin.murphy@arm.com>
Signed-off-by: Kefeng Wang <wangkefeng.wang@huawei.com>
Reviewed-by: Sergey Senozhatsky <sergey.senozhatsky@gmail.com>
Signed-off-by: Petr Mladek <pmladek@suse.com>
Based on 1 normalized pattern(s):
this program is free software you can redistribute it and or modify
it under the terms of the gnu general public license version 2 as
published by the free software foundation this program is
distributed in the hope that it will be useful but without any
warranty without even the implied warranty of merchantability or
fitness for a particular purpose see the gnu general public license
for more details you should have received a copy of the gnu general
public license along with this program if not write to the free
software foundation inc 59 temple place suite 330 boston ma 02111
1307 usa
extracted by the scancode license scanner the SPDX license identifier
GPL-2.0-only
has been chosen to replace the boilerplate/reference in 136 file(s).
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Reviewed-by: Alexios Zavras <alexios.zavras@intel.com>
Reviewed-by: Allison Randal <allison@lohutok.net>
Cc: linux-spdx@vger.kernel.org
Link: https://lkml.kernel.org/r/20190530000436.384967451@linutronix.de
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
With skip set to 1, I get a traceback like this:
[ 106.867637] DMA-API: Mapped at:
[ 106.870784] afu_dma_map_region+0x2cd/0x4f0 [dfl_afu]
[ 106.875839] afu_ioctl+0x258/0x380 [dfl_afu]
[ 106.880108] do_vfs_ioctl+0xa9/0x720
[ 106.883688] ksys_ioctl+0x60/0x90
[ 106.887007] __x64_sys_ioctl+0x16/0x20
With the previous value of 2, afu_dma_map_region was being omitted. I
suspect that the code paths have simply changed since the value of 2 was
chosen a decade ago, but it's also possible that it varies based on which
mapping function was used, compiler inlining choices, etc. In any case,
it's best to err on the side of skipping less.
Signed-off-by: Scott Wood <swood@redhat.com>
Signed-off-by: Christoph Hellwig <hch@lst.de>
While debugging a DMA mapping leak, I needed to access
debug_dma_dump_mappings() but easily from user space.
This patch adds a /sys/kernel/debug/dma-api/dump file which contain all
current DMA mapping.
Signed-off-by: Corentin Labbe <clabbe@baylibre.com>
Signed-off-by: Christoph Hellwig <hch@lst.de>
When calling debugfs functions, there is no need to ever check the
return value. The function can work or not, but the code logic should
never do something different based on this.
Also delete the variables for the file dentries for the debugfs entries
as they are never used at all once they are created.
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Reviewed-by: Robin Murphy <robin.murphy@arm.com>
[hch: moved dma_debug_dent to function scope and renamed it]
Signed-off-by: Christoph Hellwig <hch@lst.de>
Now that the slow path DMA API calls are implemented out of line a few
helpers only used by them don't need to be exported anymore.
Signed-off-by: Christoph Hellwig <hch@lst.de>
And also switch the way we implement the unmap side around to stay
consistent. This ensures dma-debug works again because it records which
function we used for mapping to ensure it is also used for unmapping,
and also reduces further code duplication. Last but not least this
also officially allows calling dma_sync_single_* for mappings created
using dma_map_page, which is perfectly fine given that the sync calls
only take a dma_addr_t, but not a virtual address or struct page.
Fixes: 7f0fee242e ("dma-mapping: merge dma_unmap_page_attrs and dma_unmap_single_attrs")
Signed-off-by: Christoph Hellwig <hch@lst.de>
Tested-by: LABBE Corentin <clabbe.montjoie@gmail.com>
We can just call the regular calls after adding offset the the address instead
of reimplementing them.
Signed-off-by: Christoph Hellwig <hch@lst.de>
Acked-by: Jesper Dangaard Brouer <brouer@redhat.com>
Tested-by: Jesper Dangaard Brouer <brouer@redhat.com>
Tested-by: Tony Luck <tony.luck@intel.com>
DMA debug entries are one of those things which aren't that useful
individually - we will always want some larger quantity of them - and
which we don't really need to manage the exact number of - we only care
about having 'enough'. In that regard, the current behaviour of creating
them one-by-one leads to a lot of unwarranted function call overhead and
memory wasted on alignment padding.
Now that we don't have to worry about freeing anything via
dma_debug_resize_entries(), we can optimise the allocation behaviour by
grabbing whole pages at once, which will save considerably on the
aforementioned overheads, and probably offer a little more cache/TLB
locality benefit for traversing the lists under normal operation. This
should also give even less reason for an architecture-level override of
the preallocation size, so make the definition unconditional - if there
is still any desire to change the compile-time value for some platforms
it would be better off as a Kconfig option anyway.
Since freeing a whole page of entries at once becomes enough of a
challenge that it's not really worth complicating dma_debug_init(), we
may as well tweak the preallocation behaviour such that as long as we
manage to allocate *some* pages, we can leave debugging enabled on a
best-effort basis rather than otherwise wasting them.
Signed-off-by: Robin Murphy <robin.murphy@arm.com>
Tested-by: Qian Cai <cai@lca.pw>
Signed-off-by: Christoph Hellwig <hch@lst.de>
With the only caller now gone, we can clean up this part of dma-debug's
exposed internals and make way to tweak the allocation behaviour.
Signed-off-by: Robin Murphy <robin.murphy@arm.com>
Tested-by: Qian Cai <cai@lca.pw>
Signed-off-by: Christoph Hellwig <hch@lst.de>
Now that we can dynamically allocate DMA debug entries to cope with
drivers maintaining excessively large numbers of live mappings, a driver
which *does* actually have a bug leaking mappings (and is not unloaded)
will no longer trigger the "DMA-API: debugging out of memory - disabling"
message until it gets to actual kernel OOM conditions, which means it
could go unnoticed for a while. To that end, let's inform the user each
time the pool has grown to a multiple of its initial size, which should
make it apparent that they either have a leak or might want to increase
the preallocation size.
Signed-off-by: Robin Murphy <robin.murphy@arm.com>
Tested-by: Qian Cai <cai@lca.pw>
Signed-off-by: Christoph Hellwig <hch@lst.de>
Certain drivers such as large multi-queue network adapters can use pools
of mapped DMA buffers larger than the default dma_debug_entry pool of
65536 entries, with the result that merely probing such a device can
cause DMA debug to disable itself during boot unless explicitly given an
appropriate "dma_debug_entries=..." option.
Developers trying to debug some other driver on such a system may not be
immediately aware of this, and at worst it can hide bugs if they fail to
realise that dma-debug has already disabled itself unexpectedly by the
time their code of interest gets to run. Even once they do realise, it
can be a bit of a pain to emprirically determine a suitable number of
preallocated entries to configure, short of massively over-allocating.
There's really no need for such a static limit, though, since we can
quite easily expand the pool at runtime in those rare cases that the
preallocated entries are insufficient, which is arguably the least
surprising and most useful behaviour. To that end, refactor the
prealloc_memory() logic a little bit to generalise it for runtime
reallocations as well.
Signed-off-by: Robin Murphy <robin.murphy@arm.com>
Tested-by: Qian Cai <cai@lca.pw>
Signed-off-by: Christoph Hellwig <hch@lst.de>
Use pr_fmt() to generate the "DMA-API: " prefix consistently. This
results in it being added to a couple of pr_*() messages which were
missing it before, and for the err_printk() calls moves it to the actual
start of the message instead of somewhere in the middle.
Signed-off-by: Robin Murphy <robin.murphy@arm.com>
Tested-by: Qian Cai <cai@lca.pw>
Signed-off-by: Christoph Hellwig <hch@lst.de>
