net_dma can cause data to be copied to a stale mapping if a
copy-on-write fault occurs during dma. The application sees missing
data.
The following trace is triggered by modifying the kernel to WARN if it
ever triggers copy-on-write on a page that is undergoing dma:
WARNING: CPU: 24 PID: 2529 at lib/dma-debug.c:485 debug_dma_assert_idle+0xd2/0x120()
ioatdma 0000:00:04.0: DMA-API: cpu touching an active dma mapped page [pfn=0x16bcd9]
Modules linked in: iTCO_wdt iTCO_vendor_support ioatdma lpc_ich pcspkr dca
CPU: 24 PID: 2529 Comm: linbug Tainted: G W 3.13.0-rc1+ #353
00000000000001e5 ffff88016f45f688 ffffffff81751041 ffff88017ab0ef70
ffff88016f45f6d8 ffff88016f45f6c8 ffffffff8104ed9c ffffffff810f3646
ffff8801768f4840 0000000000000282 ffff88016f6cca10 00007fa2bb699349
Call Trace:
[<ffffffff81751041>] dump_stack+0x46/0x58
[<ffffffff8104ed9c>] warn_slowpath_common+0x8c/0xc0
[<ffffffff810f3646>] ? ftrace_pid_func+0x26/0x30
[<ffffffff8104ee86>] warn_slowpath_fmt+0x46/0x50
[<ffffffff8139c062>] debug_dma_assert_idle+0xd2/0x120
[<ffffffff81154a40>] do_wp_page+0xd0/0x790
[<ffffffff811582ac>] handle_mm_fault+0x51c/0xde0
[<ffffffff813830b9>] ? copy_user_enhanced_fast_string+0x9/0x20
[<ffffffff8175fc2c>] __do_page_fault+0x19c/0x530
[<ffffffff8175c196>] ? _raw_spin_lock_bh+0x16/0x40
[<ffffffff810f3539>] ? trace_clock_local+0x9/0x10
[<ffffffff810fa1f4>] ? rb_reserve_next_event+0x64/0x310
[<ffffffffa0014c00>] ? ioat2_dma_prep_memcpy_lock+0x60/0x130 [ioatdma]
[<ffffffff8175ffce>] do_page_fault+0xe/0x10
[<ffffffff8175c862>] page_fault+0x22/0x30
[<ffffffff81643991>] ? __kfree_skb+0x51/0xd0
[<ffffffff813830b9>] ? copy_user_enhanced_fast_string+0x9/0x20
[<ffffffff81388ea2>] ? memcpy_toiovec+0x52/0xa0
[<ffffffff8164770f>] skb_copy_datagram_iovec+0x5f/0x2a0
[<ffffffff8169d0f4>] tcp_rcv_established+0x674/0x7f0
[<ffffffff816a68c5>] tcp_v4_do_rcv+0x2e5/0x4a0
[..]
---[ end trace e30e3b01191b7617 ]---
Mapped at:
[<ffffffff8139c169>] debug_dma_map_page+0xb9/0x160
[<ffffffff8142bf47>] dma_async_memcpy_pg_to_pg+0x127/0x210
[<ffffffff8142cce9>] dma_memcpy_pg_to_iovec+0x119/0x1f0
[<ffffffff81669d3c>] dma_skb_copy_datagram_iovec+0x11c/0x2b0
[<ffffffff8169d1ca>] tcp_rcv_established+0x74a/0x7f0:
...the problem is that the receive path falls back to cpu-copy in
several locations and this trace is just one of the areas. A few
options were considered to fix this:
1/ sync all dma whenever a cpu copy branch is taken
2/ modify the page fault handler to hold off while dma is in-flight
Option 1 adds yet more cpu overhead to an "offload" that struggles to compete
with cpu-copy. Option 2 adds checks for behavior that is already documented as
broken when using get_user_pages(). At a minimum a debug mode is warranted to
catch and flag these violations of the dma-api vs get_user_pages().
Thanks to David for his reproducer.
Cc: <stable@vger.kernel.org>
Cc: Dave Jiang <dave.jiang@intel.com>
Cc: Vinod Koul <vinod.koul@intel.com>
Cc: Alexander Duyck <alexander.h.duyck@intel.com>
Reported-by: David Whipple <whipple@securedatainnovations.ch>
Acked-by: David S. Miller <davem@davemloft.net>
Signed-off-by: Dan Williams <dan.j.williams@intel.com>
When a channel fails to initialise, we error out and clean up any
previously unregistered channels by walking the entire xordev->channels
array. Unfortunately, there are paths which end up storing an error
pointer in this array, which we then try and dereference in the cleanup
code, which causes an oops.
Fix this by avoiding writing invalid pointers to this array in the first
place.
Tested-by: Aaro Koskinen <aaro.koskinen@iki.fi>
Signed-off-by: Russell King <rmk+kernel@arm.linux.org.uk>
Signed-off-by: Dan Williams <dan.j.williams@intel.com>
The driver-specific unmap code was removed in:
commit 54f8d501e8
Author: Bartlomiej Zolnierkiewicz <b.zolnierkie@samsung.com>
Date: Fri Oct 18 19:35:32 2013 +0200
dmaengine: remove DMA unmap from drivers
which had the side-effect of not unmapping the self-test mappings.
Fix this by using dmaengine_unmap_data in the self-test routines.
In addition, since dmaengine_unmap() assumes that all mappings were created
with dma_map_page, this commit changes the single mapping to a page mapping
to avoid an incorrect unmapping of the memcpy self-test.
The allocation could be changed to be alloc_page(), but sticking to kmalloc
results in a less intrusive patch. The size of the test buffer is increased,
since dma_map_page() seem to fail when the source and destination pages are
the same page.
Signed-off-by: Ezequiel Garcia <ezequiel.garcia@free-electrons.com>
Signed-off-by: Dan Williams <dan.j.williams@intel.com>
The higher order mempools support raid operations, and we want to
disable them when raid support is not enabled. Making them conditional
on ASYNC_TX_DMA is not sufficient as other users (specifically dmatest)
will also issue raid operations. Make raid drivers explicitly request
that the core carry the higher order pools.
Reported-by: Ezequiel Garcia <ezequiel.garcia@free-electrons.com>
Tested-by: Ezequiel Garcia <ezequiel.garcia@free-electrons.com>
Signed-off-by: Dan Williams <dan.j.williams@intel.com>
drivers/dma/dmatest.c:543:11: warning: passing argument 1 of 'virt_to_phys' makes pointer from integer without a cast [enabled by default]
mips expects virt_to_phys() to take a pointer. Fix up the types accordingly.
Signed-off-by: Dan Williams <dan.j.williams@intel.com>
drivers/dma/fsldma.c: In function 'fsldma_cleanup_descriptor':
drivers/dma/fsldma.c:860:6: warning: unused variable 'len' [-Wunused-variable]
drivers/dma/fsldma.c:859:13: warning: unused variable 'dst' [-Wunused-variable]
drivers/dma/fsldma.c:858:13: warning: unused variable 'src' [-Wunused-variable]
drivers/dma/fsldma.c:857:17: warning: unused variable 'dev' [-Wunused-variable]
- due to unmap changes
drivers/dma/fsldma.c: In function 'fsl_dma_tx_submit':
drivers/dma/fsldma.c:428:2: warning: 'cookie' may be used uninitialized in this function [-Wuninitialized]
- long standing warning
Cc: Bartlomiej Zolnierkiewicz <b.zolnierkie@samsung.com>
Cc: Li Yang <leoli@freescale.com>
Cc: Zhang Wei <zw@zh-kernel.org>
Signed-off-by: Dan Williams <dan.j.williams@intel.com>
drivers/dma/ppc4xx/adma.c:1507:6: warning: unused variable 'i' [-Wunused-variable]
- due to unmap reworks
drivers/dma/ppc4xx/adma.c:3900:2: warning: format '%s' expects a matching 'char *' argument [-Wformat]
- due to memset removal
drivers/dma/ppc4xx/adma.c:538:13: warning: 'ppc440spe_desc_init_memset' defined but not used [-Wunused-function]
- due to memset removal
Cc: Bartlomiej Zolnierkiewicz <b.zolnierkie@samsung.com>
Signed-off-by: Dan Williams <dan.j.williams@intel.com>
commit 54f8d501e8 ('dmaengine: remove DMA unmap from drivers')
refactored some code which resulted in an unused function in the at_hdmac
driver:
drivers/dma/at_hdmac_regs.h:350:23: warning: 'chan2parent' defined but
not used [-Wunused-function]
Fixes: 54f8d501e8 ('dmaengine: remove DMA unmap from drivers')
Signed-off-by: Olof Johansson <olof@lixom.net>
Cc: Bartlomiej Zolnierkiewicz <b.zolnierkie@samsung.com>
Acked-by: Nicolas Ferre <nicolas.ferre@atmel.com>
Acked-by: Bartlomiej Zolnierkiewicz <b.zolnierkie@samsung.com>
Acked-by: Vinod Koul <vinod.koul@Intel.com>
Signed-off-by: Dan Williams <dan.j.williams@intel.com>
The following commit:
54f8d501e8 dmaengine: remove DMA unmap from drivers
removed the last caller to mv_desc_get_dest_addr(), creating the
warning:
drivers/dma/mv_xor.c:57:12: warning: mv_desc_get_dest_addr defined
but not used [-Wunused-function]
Remove it.
Signed-off-by: Jason Cooper <jason@lakedaemon.net>
Acked-by: Vinod Koul <vinod.koul@intel.com>
Signed-off-by: Dan Williams <dan.j.williams@intel.com>
This driver missed the dma unmap conversion. Replace
s3c24xx_dma_unmap_buffers with dma_descriptor_unmap.
Signed-off-by: Dan Williams <dan.j.williams@intel.com>
Signed-off-by: Vinod Koul <vinod.koul@intel.com>
drivers/dma/amba-pl08x.c: In function 'pl08x_desc_free':
drivers/dma/amba-pl08x.c:1173:2: warning: passing argument 1 of 'dma_descriptor_unmap' from incompatible pointer type
include/linux/dmaengine.h:476:91: note: expected 'struct dma_async_tx_descriptor *' but argument is of type 'struct pl08x_txd *'
Fixes: d38a8c622a ("dmaengine: prepare for generic 'unmap' data")
Signed-off-by: Russell King <rmk+kernel@arm.linux.org.uk>
Signed-off-by: Dan Williams <dan.j.williams@intel.com>
Signed-off-by: Vinod Koul <vinod.koul@intel.com>
Plane information should be initialized when halted.
It may restart from the wrong plane without this patch.
Reviewed-by: Max Filippov <max.filippov@cogentembedded.com>
Signed-off-by: Kuninori Morimoto <kuninori.morimoto.gx@renesas.com>
Acked-by: Simon Horman <horms+renesas@verge.net.au>
Signed-off-by: Vinod Koul <vinod.koul@intel.com>
The device busy check method is different between
single and double planes.
It will always return "busy" without this patch
if channel used as double plane.
Reviewed-by: Max Filippov <max.filippov@cogentembedded.com>
Signed-off-by: Kuninori Morimoto <kuninori.morimoto.gx@renesas.com>
Acked-by: Simon Horman <horms+renesas@verge.net.au>
Signed-off-by: Vinod Koul <vinod.koul@intel.com>
shdma_chan_probe() can set max transfer size,
but it will be PAGE_SIZE with out this patch.
Reviewed-by: Max Filippov <max.filippov@cogentembedded.com>
Signed-off-by: Kuninori Morimoto <kuninori.morimoto.gx@renesas.com>
Acked-by: Simon Horman <horms+renesas@verge.net.au>
Signed-off-by: Vinod Koul <vinod.koul@intel.com>
Use the recently introduced DMA_COMPLETE instead of DMA_SUCCESS.
Without this patch we get the following build error:
drivers/dma/s3c24xx-dma.c: In function ‘s3c24xx_dma_tx_status’:
drivers/dma/s3c24xx-dma.c:798:13: error: ‘DMA_SUCCESS’ undeclared
(first use in this function)
Signed-off-by: Sachin Kamat <sachin.kamat@linaro.org>
Acked-by: Heiko Stuebner <heiko@sntech.de>
Acked-by: Kukjin Kim <kgene@kernel.org>
Signed-off-by: Vinod Koul <vinod.koul@intel.com>
Pull slave-dmaengine changes from Vinod Koul:
"This brings for slave dmaengine:
- Change dma notification flag to DMA_COMPLETE from DMA_SUCCESS as
dmaengine can only transfer and not verify validaty of dma
transfers
- Bunch of fixes across drivers:
- cppi41 driver fixes from Daniel
- 8 channel freescale dma engine support and updated bindings from
Hongbo
- msx-dma fixes and cleanup by Markus
- DMAengine updates from Dan:
- Bartlomiej and Dan finalized a rework of the dma address unmap
implementation.
- In the course of testing 1/ a collection of enhancements to
dmatest fell out. Notably basic performance statistics, and
fixed / enhanced test control through new module parameters
'run', 'wait', 'noverify', and 'verbose'. Thanks to Andriy and
Linus [Walleij] for their review.
- Testing the raid related corner cases of 1/ triggered bugs in
the recently added 16-source operation support in the ioatdma
driver.
- Some minor fixes / cleanups to mv_xor and ioatdma"
* 'next' of git://git.infradead.org/users/vkoul/slave-dma: (99 commits)
dma: mv_xor: Fix mis-usage of mmio 'base' and 'high_base' registers
dma: mv_xor: Remove unneeded NULL address check
ioat: fix ioat3_irq_reinit
ioat: kill msix_single_vector support
raid6test: add new corner case for ioatdma driver
ioatdma: clean up sed pool kmem_cache
ioatdma: fix selection of 16 vs 8 source path
ioatdma: fix sed pool selection
ioatdma: Fix bug in selftest after removal of DMA_MEMSET.
dmatest: verbose mode
dmatest: convert to dmaengine_unmap_data
dmatest: add a 'wait' parameter
dmatest: add basic performance metrics
dmatest: add support for skipping verification and random data setup
dmatest: use pseudo random numbers
dmatest: support xor-only, or pq-only channels in tests
dmatest: restore ability to start test at module load and init
dmatest: cleanup redundant "dmatest: " prefixes
dmatest: replace stored results mechanism, with uniform messages
Revert "dmatest: append verify result to results"
...
Pull dmaengine changes from Dan
1/ Bartlomiej and Dan finalized a rework of the dma address unmap
implementation.
2/ In the course of testing 1/ a collection of enhancements to dmatest
fell out. Notably basic performance statistics, and fixed / enhanced
test control through new module parameters 'run', 'wait', 'noverify',
and 'verbose'. Thanks to Andriy and Linus for their review.
3/ Testing the raid related corner cases of 1/ triggered bugs in the
recently added 16-source operation support in the ioatdma driver.
4/ Some minor fixes / cleanups to mv_xor and ioatdma.
Conflicts:
drivers/dma/dmatest.c
Signed-off-by: Vinod Koul <vinod.koul@intel.com>
Pull trivial tree updates from Jiri Kosina:
"Usual earth-shaking, news-breaking, rocket science pile from
trivial.git"
* 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/jikos/trivial: (23 commits)
doc: usb: Fix typo in Documentation/usb/gadget_configs.txt
doc: add missing files to timers/00-INDEX
timekeeping: Fix some trivial typos in comments
mm: Fix some trivial typos in comments
irq: Fix some trivial typos in comments
NUMA: fix typos in Kconfig help text
mm: update 00-INDEX
doc: Documentation/DMA-attributes.txt fix typo
DRM: comment: `halve' -> `half'
Docs: Kconfig: `devlopers' -> `developers'
doc: typo on word accounting in kprobes.c in mutliple architectures
treewide: fix "usefull" typo
treewide: fix "distingush" typo
mm/Kconfig: Grammar s/an/a/
kexec: Typo s/the/then/
Documentation/kvm: Update cpuid documentation for steal time and pv eoi
treewide: Fix common typo in "identify"
__page_to_pfn: Fix typo in comment
Correct some typos for word frequency
clk: fixed-factor: Fix a trivial typo
...
Despite requesting two memory resources, called 'base' and 'high_base', the
driver uses explicitly only the former. The latter is being used implicitly
by addressing at offset +0x200, which in practice accesses high_base.
In other words, the current driver breaks if the second memory resource
is ever place at an offset different from +0x200.
This patch fixes the above by defining the registers with the offset from
high_base, and use high_base explicitly where appropriate.
Signed-off-by: Ezequiel Garcia <ezequiel.garcia@free-electrons.com>
Acked-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
Signed-off-by: Dan Williams <dan.j.williams@intel.com>
This mmio address is checked at probe-time, which makes this test
redundant. Let's just remove it.
Acked-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
Signed-off-by: Ezequiel Garcia <ezequiel.garcia@free-electrons.com>
Signed-off-by: Dan Williams <dan.j.williams@intel.com>
The implementation of ioat3_irq_reinit has two bugs:
1/ The mode is incorrectly set to MSIX for the MSI case
2/ The 'dev_id' parameter to free_irq is the ioatdma_device not the channel in
the msi and intx case
Include a small cleanup to clarify that ioat3_irq_reinit is only for bwd
hardware
Cc: Dave Jiang <dave.jiang@intel.com>
Signed-off-by: Dan Williams <dan.j.williams@intel.com>
Once we have determined that we will not have all of our desired msix
vectors there is no point in attempting a single msix allocation. The
driver will already need to read registers to determine the source of
the interrupt the fact that it is msix is moot. Fallback directly to
msi.
Reported-by: Alexander Gordeev <agordeev@redhat.com>
Cc: Dave Jiang <dave.jiang@intel.com>
Signed-off-by: Dan Williams <dan.j.williams@intel.com>
Use a single cache for all sed allocations. No need to make it per
channel. This also avoids the slub_debug warnings for multiple caches
with the same name.
Switching to dmam_pool_create() to fix leaking the dma pools on
initialization failure and lets us kill ioat3_dma_remove().
Cc: Dave Jiang <dave.jiang@intel.com>
Acked-by: Dave Jiang <dave.jiang@intel.com>
Signed-off-by: Dan Williams <dan.j.williams@intel.com>
When performing continuations there are implied sources that need to be
added to the source count. Quoting dma_set_maxpq:
/* dma_maxpq - reduce maxpq in the face of continued operations
* @dma - dma device with PQ capability
* @flags - to check if DMA_PREP_CONTINUE and DMA_PREP_PQ_DISABLE_P are set
*
* When an engine does not support native continuation we need 3 extra
* source slots to reuse P and Q with the following coefficients:
* 1/ {00} * P : remove P from Q', but use it as a source for P'
* 2/ {01} * Q : use Q to continue Q' calculation
* 3/ {00} * Q : subtract Q from P' to cancel (2)
*
* In the case where P is disabled we only need 1 extra source:
* 1/ {01} * Q : use Q to continue Q' calculation
*/
...fix the selection of the 16 source path to take these implied sources
into account.
Note this also kills the BUG_ON(src_cnt < 9) check in
__ioat3_prep_pq16_lock(). Besides not accounting for implied sources
the check is redundant given we already made the path selection.
Cc: <stable@vger.kernel.org>
Cc: Dave Jiang <dave.jiang@intel.com>
Acked-by: Dave Jiang <dave.jiang@intel.com>
Signed-off-by: Dan Williams <dan.j.williams@intel.com>
The array to lookup the sed pool based on the number of sources
(pq16_idx_to_sedi) is 16 entries and expects a max source index.
However, we pass the total source count which runs off the end of the
array when src_cnt == 16. The minimal fix is to just pass src_cnt-1,
but given we know the source count is > 8 we can just calculate the sed
pool by (src_cnt - 2) >> 3.
Cc: Dave Jiang <dave.jiang@intel.com>
Cc: <stable@vger.kernel.org>
Acked-by: Dave Jiang <dave.jiang@intel.com>
Signed-off-by: Dan Williams <dan.j.williams@intel.com>
Commit 48a9db4 (3.11) removed the memset op in the xor selftest for ioatdma.
The issue is that with the removal of that op, it never replaced the memset
with a CPU memset. The memory being operated on is expected to be zeroes but
was not. This is causing the xor selftest to fail.
Cc: <stable@vger.kernel.org>
Signed-off-by: Dave Jiang <dave.jiang@intel.com>
Signed-off-by: Dan Williams <dan.j.williams@intel.com>
Verbose mode turns on test success messages, by default we only output
test summaries and failure results.
Also cleaned up some stray quotes, leftover from putting the result
message format string all on one line.
Cc: Andy Shevchenko <andriy.shevchenko@linux.intel.com>
Signed-off-by: Dan Williams <dan.j.williams@intel.com>
Remove the open coded unmap and add coverage for this core functionality
to dmatest.
Also fixes up a couple places where we leaked dma mappings.
Signed-off-by: Dan Williams <dan.j.williams@intel.com>
Allows for scripting test runs by module load / unload. Prevent module
load from returning until 'iterations' (finite) tests have completed, or
cause reads of the 'wait' parameter in sysfs to pause until the tests
are done.
Also killed the local waitqueue since we can just let the thread exit
naturally as long as we hold a reference.
Cc: Nicolas Ferre <nicolas.ferre@atmel.com>
Signed-off-by: Dan Williams <dan.j.williams@intel.com>
Add iops and throughput to the summary output.
Acked-by: Andy Shevchenko <andriy.shevchenko@linux.intel.com>
Signed-off-by: Dan Williams <dan.j.williams@intel.com>
Towards enabling dmatest to checkout performance add a 'noverify' mode.
Acked-by: Andy Shevchenko <andriy.shevchenko@linux.intel.com>
Signed-off-by: Dan Williams <dan.j.williams@intel.com>
There is no need for dmatest to drain the entropy pool.
It would be nice to one day have repeatable runs, but would need a
larger rework to synchronize and order calls to the rng across test
threads.
Acked-by: Andy Shevchenko <andriy.shevchenko@linux.intel.com>
Signed-off-by: Dan Williams <dan.j.williams@intel.com>
Currently we only test raid channels that happen to also have 'copy'
capability. Search for capable channels that do not have DMA_MEMCPY.
Note the return value from run_threaded_test never really made sense
because it could return errors after successfully starting tests. We
already have the test results per channel so missing channels can be
detected at that time.
Acked-by: Andy Shevchenko <andriy.shevchenko@linux.intel.com>
Signed-off-by: Dan Williams <dan.j.williams@intel.com>
1/ move 'run' control to a module parameter so we can do:
modprobe dmatest run=1. With this moved the rest of the debugfs
boilerplate can go.
2/ Fix parameter initialization. Previously the test was being started
without taking the parameters into account in the built-in case.
Also killed off the '__' version of some routines. The new rule is just
hold the lock when calling a *threaded_test() routine.
Acked-by: Linus Walleij <linus.walleij@linaro.org>
Cc: Andy Shevchenko <andriy.shevchenko@linux.intel.com>
Acked-by: Andy Shevchenko <andriy.shevchenko@linux.intel.com>
Signed-off-by: Dan Williams <dan.j.williams@intel.com>
...now that we have a common pr_fmt.
Acked-by: Andy Shevchenko <andriy.shevchenko@linux.intel.com>
Signed-off-by: Dan Williams <dan.j.williams@intel.com>
For long running tests the tracking results in a memory leak for the "ok"
results, and for the failures the kernel log should be sufficient. Provide a
uniform format for error messages so they can be easily parsed and remove the
debugfs file.
Cc: Andy Shevchenko <andriy.shevchenko@linux.intel.com>
Signed-off-by: Dan Williams <dan.j.williams@intel.com>
This reverts commit d86b2f298e.
The kernel log buffer is sufficient for collecting test results. The
current logging OOMs the machine on long running tests, and usually only
the first error is relevant. It is better to stop on error and parse
the kernel output. If output volume becomes an issue we can always
investigate using trace messages.
Cc: Andy Shevchenko <andriy.shevchenko@linux.intel.com>
Acked-by: Andy Shevchenko <andriy.shevchenko@linux.intel.com>
Signed-off-by: Dan Williams <dan.j.williams@intel.com>
Remove support for DMA unmapping from drivers as it is no longer
needed (DMA core code is now handling it).
Cc: Vinod Koul <vinod.koul@intel.com>
Cc: Tomasz Figa <t.figa@samsung.com>
Cc: Dave Jiang <dave.jiang@intel.com>
Signed-off-by: Bartlomiej Zolnierkiewicz <b.zolnierkie@samsung.com>
Signed-off-by: Kyungmin Park <kyungmin.park@samsung.com>
[djbw: fix up chan2parent() unused warning in drivers/dma/dw/core.c]
Signed-off-by: Dan Williams <dan.j.williams@intel.com>
Use the generic unmap object to unmap dma buffers.
Cc: Vinod Koul <vinod.koul@intel.com>
Cc: Tomasz Figa <t.figa@samsung.com>
Cc: Dave Jiang <dave.jiang@intel.com>
Reported-by: Bartlomiej Zolnierkiewicz <b.zolnierkie@samsung.com>
[bzolnier: keep temporary dma_dest array in do_async_gen_syndrome()]
Signed-off-by: Bartlomiej Zolnierkiewicz <b.zolnierkie@samsung.com>
Signed-off-by: Kyungmin Park <kyungmin.park@samsung.com>
Signed-off-by: Dan Williams <dan.j.williams@intel.com>
Hang a common 'unmap' object off of dma descriptors for the purpose of
providing a unified unmapping interface. The lifetime of a mapping may
span multiple descriptors, so these unmap objects are reference counted
by related descriptor.
Cc: Vinod Koul <vinod.koul@intel.com>
Cc: Tomasz Figa <t.figa@samsung.com>
Cc: Dave Jiang <dave.jiang@intel.com>
[bzolnier: fix IS_ENABLED() check]
[bzolnier: fix release ordering in dmaengine_destroy_unmap_pool()]
[bzolnier: fix check for success in dmaengine_init_unmap_pool()]
[bzolnier: use mempool_free() instead of kmem_cache_free()]
[bzolnier: add missing unmap->len initializations]
[bzolnier: add __init tag to dmaengine_init_unmap_pool()]
Signed-off-by: Bartlomiej Zolnierkiewicz <b.zolnierkie@samsung.com>
Signed-off-by: Kyungmin Park <kyungmin.park@samsung.com>
[djbw: move DMAENGINE=n support to this patch for async_tx]
Signed-off-by: Dan Williams <dan.j.williams@intel.com>
Add a hook for a common dma unmap implementation to enable removal of
the per driver custom unmap code. (A reworked version of Bartlomiej
Zolnierkiewicz's patches to remove the custom callbacks and the size
increase of dma_async_tx_descriptor for drivers that don't care about
raid).
Cc: Vinod Koul <vinod.koul@intel.com>
Cc: Tomasz Figa <t.figa@samsung.com>
Cc: Dave Jiang <dave.jiang@intel.com>
[bzolnier: prepare pl330 driver for adding missing unmap while at it]
Signed-off-by: Bartlomiej Zolnierkiewicz <b.zolnierkie@samsung.com>
Signed-off-by: Kyungmin Park <kyungmin.park@samsung.com>
Signed-off-by: Dan Williams <dan.j.williams@intel.com>
Copying from page to page (dma_async_memcpy_pg_to_pg) is the superset,
make the other two apis use that one in preparation for providing a
common dma unmap implementation. The common implementation just wants
to assume all buffers are mapped with dma_map_page().
Cc: Vinod Koul <vinod.koul@intel.com>
Cc: Tomasz Figa <t.figa@samsung.com>
Cc: Dave Jiang <dave.jiang@intel.com>
Signed-off-by: Bartlomiej Zolnierkiewicz <b.zolnierkie@samsung.com>
Signed-off-by: Kyungmin Park <kyungmin.park@samsung.com>
Signed-off-by: Dan Williams <dan.j.williams@intel.com>
Pull DMA mask updates from Russell King:
"This series cleans up the handling of DMA masks in a lot of drivers,
fixing some bugs as we go.
Some of the more serious errors include:
- drivers which only set their coherent DMA mask if the attempt to
set the streaming mask fails.
- drivers which test for a NULL dma mask pointer, and then set the
dma mask pointer to a location in their module .data section -
which will cause problems if the module is reloaded.
To counter these, I have introduced two helper functions:
- dma_set_mask_and_coherent() takes care of setting both the
streaming and coherent masks at the same time, with the correct
error handling as specified by the API.
- dma_coerce_mask_and_coherent() which resolves the problem of
drivers forcefully setting DMA masks. This is more a marker for
future work to further clean these locations up - the code which
creates the devices really should be initialising these, but to fix
that in one go along with this change could potentially be very
disruptive.
The last thing this series does is prise away some of Linux's addition
to "DMA addresses are physical addresses and RAM always starts at
zero". We have ARM LPAE systems where all system memory is above 4GB
physical, hence having DMA masks interpreted by (eg) the block layers
as describing physical addresses in the range 0..DMAMASK fails on
these platforms. Santosh Shilimkar addresses this in this series; the
patches were copied to the appropriate people multiple times but were
ignored.
Fixing this also gets rid of some ARM weirdness in the setup of the
max*pfn variables, and brings ARM into line with every other Linux
architecture as far as those go"
* 'for-linus-dma-masks' of git://git.linaro.org/people/rmk/linux-arm: (52 commits)
ARM: 7805/1: mm: change max*pfn to include the physical offset of memory
ARM: 7797/1: mmc: Use dma_max_pfn(dev) helper for bounce_limit calculations
ARM: 7796/1: scsi: Use dma_max_pfn(dev) helper for bounce_limit calculations
ARM: 7795/1: mm: dma-mapping: Add dma_max_pfn(dev) helper function
ARM: 7794/1: block: Rename parameter dma_mask to max_addr for blk_queue_bounce_limit()
ARM: DMA-API: better handing of DMA masks for coherent allocations
ARM: 7857/1: dma: imx-sdma: setup dma mask
DMA-API: firmware/google/gsmi.c: avoid direct access to DMA masks
DMA-API: dcdbas: update DMA mask handing
DMA-API: dma: edma.c: no need to explicitly initialize DMA masks
DMA-API: usb: musb: use platform_device_register_full() to avoid directly messing with dma masks
DMA-API: crypto: remove last references to 'static struct device *dev'
DMA-API: crypto: fix ixp4xx crypto platform device support
DMA-API: others: use dma_set_coherent_mask()
DMA-API: staging: use dma_set_coherent_mask()
DMA-API: usb: use new dma_coerce_mask_and_coherent()
DMA-API: usb: use dma_set_coherent_mask()
DMA-API: parport: parport_pc.c: use dma_coerce_mask_and_coherent()
DMA-API: net: octeon: use dma_coerce_mask_and_coherent()
DMA-API: net: nxp/lpc_eth: use dma_coerce_mask_and_coherent()
...
