Currently the handler ignores the channel 0 interrupt and thus doesn't ack
it properly. This is done in order to allow sdma_run_channel0() to poll
on the irq status bit, as this function may be called in atomic context,
but needs to know when the channel has finished.
This works mostly, as the polling happens under a spinlock, disabling IRQs
on the local CPU, leaving only a very slight race window for a spurious
IRQ to happen if the handler is executed on another CPU in an SMP system.
Still this is clearly suboptimal.
This behavior turns into a real problem on an RT system, where the spinlock
doesn't disable IRQs on the local CPU. Not acking the IRQ in the handler
in such a setup is very likely to drown the CPU in an IRQ storm, leaving
it unable to make any progress in the polling loop, leading to the IRQ
never being acked.
Fix this by properly acknowledging the channel 0 IRQ in the handler.
As the IRQ status bit can no longer be used to poll for the channel
completion, switch over to using the SDMA_H_STATSTOP register for this
purpose, where bit 0 is cleared by the hardware when the channel is done.
Signed-off-by: Michael Olbrich <m.olbrich@pengutronix.de>
Signed-off-by: Lucas Stach <l.stach@pengutronix.de>
Signed-off-by: Vinod Koul <vinod.koul@intel.com>
In case of error, the function platform_device_register_full()
returns ERR_PTR() and never returns NULL. The NULL test in the
return value check should be replaced with IS_ERR().
Signed-off-by: Wei Yongjun <yongjun_wei@trendmicro.com.cn>
Signed-off-by: Vinod Koul <vinod.koul@intel.com>
The new mv_xor_v2 driver supports the XOR engines found in the 64-bits
ARM from Marvell of the Armada 7K and Armada 8K family. This XOR
engine is a completely new hardware block, entirely different from the
one used on previous Marvell Armada platforms, which use the existing
mv_xor driver.
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
Signed-off-by: Vinod Koul <vinod.koul@intel.com>
This commit adds the Device Tree binding documentation for the Marvell
XOR v2 engine, which is found on Marvell Armada 7K/8K ARM64 SoCs.
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
Acked-by: Rob Herring <robh@kernel.org>
Signed-off-by: Vinod Koul <vinod.koul@intel.com>
The documentation states one should make sure both DE and TE are cleared
before starting a transaction. This patch extends the current warning to
look at both DE and TE.
Based on previous work from Muhammad Hamza Farooq.
Suggested-by: Muhammad Hamza Farooq <mfarooq@visteon.com>
Signed-off-by: Niklas Söderlund <niklas.soderlund+renesas@ragnatech.se>
Acked-by: Laurent Pinchart <laurent.pinchart@ideasonboard.com>
Signed-off-by: Vinod Koul <vinod.koul@intel.com>
The hardware might have complete the transfer but the interrupt handler
might not have had a chance to run. If rcar_dmac_chan_get_residue()
which reads HW registers finds that there is no residue return
DMA_COMPLETE.
Signed-off-by: Muhammad Hamza Farooq <mfarooq@visteon.com>
Signed-off-by: Geert Uytterhoeven <geert+renesas@glider.be>
[Niklas: add explanation in commit message]
Signed-off-by: Niklas Söderlund <niklas.soderlund+renesas@ragnatech.se>
Acked-by: Laurent Pinchart <laurent.pinchart@ideasonboard.com>
Signed-off-by: Vinod Koul <vinod.koul@intel.com>
The bam_dma driver gained runtime PM support, but that causes build
warnings whenever CONFIG_PM is disabled:
drivers/dma/qcom/bam_dma.c:1324:12: error: 'bam_dma_runtime_resume' defined but not used [-Werror=unused-function]
static int bam_dma_runtime_resume(struct device *dev)
^~~~~~~~~~~~~~~~~~~~~~
drivers/dma/qcom/bam_dma.c:1315:12: error: 'bam_dma_runtime_suspend' defined but not used [-Werror=unused-function]
static int bam_dma_runtime_suspend(struct device *dev)
This removes the incomplete #ifdef guard and instead marks all
four PM functions as __maybe_unused, which avoids this kind of
warning.
Signed-off-by: Arnd Bergmann <arnd@arndb.de>
Fixes: 7d2545599f ("dmaengine: qcom-bam-dma: Add pm_runtime support")
Signed-off-by: Vinod Koul <vinod.koul@intel.com>
When building this driver on arm64, we get a harmless type
mismatch warning:
drivers/dma/bcm2835-dma.c: In function 'bcm2835_dma_fill_cb_chain_with_sg':
include/linux/kernel.h:743:17: warning: comparison of distinct pointer types lacks a cast
(void) (&_min1 == &_min2); \
^
drivers/dma/bcm2835-dma.c:409:21: note: in expansion of macro 'min'
cb->cb->length = min(len, max_len);
This changes the type of the 'len' variable to size_t, which
avoids the problem.
Signed-off-by: Arnd Bergmann <arnd@arndb.de>
Fixes: 388cc7a281 ("dmaengine: bcm2835: add slave_sg support to bcm2835-dma")
Signed-off-by: Vinod Koul <vinod.koul@intel.com>
Return IRQ_NONE in the interrupt handler when it is called but no IRQs are
pending. This allows the system to recover in case of an interrupt storm
e.g. due to a wrong interrupt configuration setup.
Signed-off-by: Lars-Peter Clausen <lars@metafoo.de>
Signed-off-by: Vinod Koul <vinod.koul@intel.com>
Propagate errors returned by platform_get_irq() to the driver core. This
will enable proper probe deferring for the driver in case the IRQ provider
has not been registered yet.
Signed-off-by: Lars-Peter Clausen <lars@metafoo.de>
Signed-off-by: Vinod Koul <vinod.koul@intel.com>
Add MODULE_DEVICE_TABLE() for the axi-dmac driver. This allows the driver
to be loaded on demand when built as a module.
Signed-off-by: Lars-Peter Clausen <lars@metafoo.de>
Signed-off-by: Vinod Koul <vinod.koul@intel.com>
When building this driver on arm64, we get a harmless type
mismatch warning:
drivers/dma/bcm2835-dma.c: In function 'bcm2835_dma_fill_cb_chain_with_sg':
include/linux/kernel.h:743:17: warning: comparison of distinct pointer types lacks a cast
(void) (&_min1 == &_min2); \
^
drivers/dma/bcm2835-dma.c:409:21: note: in expansion of macro 'min'
cb->cb->length = min(len, max_len);
This changes the type of the 'len' variable to size_t, which
avoids the problem.
Signed-off-by: Arnd Bergmann <arnd@arndb.de>
Fixes: 388cc7a281 ("dmaengine: bcm2835: add slave_sg support to bcm2835-dma")
Signed-off-by: Vinod Koul <vinod.koul@intel.com>
Adds pm_runtime support for BAM DMA so that clock is enabled only
when there is a transaction going on to help save power.
Signed-off-by: Pramod Gurav <pramod.gurav@linaro.org>
Signed-off-by: Vinod Koul <vinod.koul@intel.com>
Commit 71f7e6cc55 ('dmaengine: tegra20-apb-dma: Only calculate residue
if txstate exists') changed the tegra_dma_tx_status() function to only
calculate the residue if there is a valid 'txstate' pointer for storing
the residue. Although this makes sense, this changed the behaviour of
the function tegra_dma_tx_status() such that if the pointer 'txstate' is
not valid, then we will return whatever state is returned by
dma_cookie_status() and no longer return the state by looking up the DMA
descriptor and returning it's state.
Please note that dma_cookie_status() will either return DMA_COMPLETE or
DMA_IN_PROGRESS. However, if dma_cookie_status() returns DMA_IN_PROGRESS
the actual status could be DMA_ERROR which will only be seen from
checking the descriptor status. Therefore, even if 'txstate' is not
valid, still check to see if there is a valid descriptor for the cookie
in question and if so return the descriptor state. Finally, ensure the
residue is still not calculated if the 'txstate' is not valid.
Signed-off-by: Jon Hunter <jonathanh@nvidia.com>
Signed-off-by: Vinod Koul <vinod.koul@intel.com>
The calculation of the DMA residue for the Tegra APB DMA is duplicated
in two places in the tegra_dma_tx_status() function. Remove this
duplicated code by moving calculation to the end of the function and
only calculating if we found a valid descriptor.
Signed-off-by: Jon Hunter <jonathanh@nvidia.com>
Signed-off-by: Vinod Koul <vinod.koul@intel.com>
Correct the grammar in the debug message when no descriptor is found.
Signed-off-by: Jon Hunter <jonathanh@nvidia.com>
Signed-off-by: Vinod Koul <vinod.koul@intel.com>
There is no point calculating the residue if there is
no txstate to store the value.
Signed-off-by: Peter Griffin <peter.griffin@linaro.org>
Signed-off-by: Vinod Koul <vinod.koul@intel.com>
There is no point in calculating the residue if state does not
exist to store the value.
Signed-off-by: Peter Griffin <peter.griffin@linaro.org>
Signed-off-by: Vinod Koul <vinod.koul@intel.com>
Currently fsl-edma doesn't clk_disable_unprepare()
its clocks on error conditions. This patch adds a
fsl_disable_clocks helper for this, and also only
disables clocks which were enabled if encountering
an error whilst enabling clocks.
Signed-off-by: Peter Griffin <peter.griffin@linaro.org>
Signed-off-by: Vinod Koul <vinod.koul@intel.com>
The bcm2835_dma_prep_dma_memcpy() function is not exported
outside the driver, so make it static to avoid the following
warning:
drivers/dma/bcm2835-dma.c:616:32: warning: symbol 'bcm2835_dma_prep_dma_memcpy' was not declared. Should it be static?
Signed-off-by: Ben Dooks <ben.dooks@codethink.co.uk>
Reviewed-by: Eric Anholt <eric@anholt.net>
Signed-off-by: Vinod Koul <vinod.koul@intel.com>
The at_xdmac_init_used_desc() and at_xdmac_prep_dma_memset()
functions are not exported outside the driver, so make them
static to avoid the following warnings:
drivers/dma/at_xdmac.c:459:6: warning: symbol 'at_xdmac_init_used_desc' was not declared. Should it be static?
drivers/dma/at_xdmac.c:1205:32: warning: symbol 'at_xdmac_prep_dma_memset' was not declared. Should it be static?
Signed-off-by: Ben Dooks <ben.dooks@codethink.co.uk>
Signed-off-by: Vinod Koul <vinod.koul@intel.com>
The driver limits the physical number of paRAM slots to be used by channels.
If the transfer needs more slots (more SGs) then the transfer is broken up
to smaller chunks. When the chunk is finished the driver will rewrite the
physical slots and continues the transfer. This set up time can take some
time and we might miss DMA events. If the intermediate set completion is
using early completion (the interrupt will happen when the last slot is
issued to the TPTC and not when the transfer is finished by the TPTC) we
will have a bit more time to update the paRAM slots and less likely to have
missed events.
Signed-off-by: Peter Ujfalusi <peter.ujfalusi@ti.com>
Signed-off-by: Vinod Koul <vinod.koul@intel.com>
For each descriptor, in addition to the memory used by the descriptors
structure itself, the driver allocates a list of chunks as well as a
buffer for hardware descriptors. Descriptors themselves are preallocated,
and allocation of the chunks and buffer is performed the first time the
descriptor is used. The memory isn't freed when the transfer is completed,
as the chunks and buffer will be needed again when the descriptor is
reused internally, so the driver keeps the memory around.
If only a few descriptors are used concurrently, the current
list_add_tail() implementation will result in all preallocated descriptors
being used before going back to the first one, and will thus allocate
chunks and a buffer for all preallocated descriptors. Using list_add()
will put the complete descriptor at the head of the list of available
descriptors, so the next transfer will be more likely to reuse a
descriptor that already has associated memory instead of one that has
never been used before.
Signed-off-by: Kuninori Morimoto <kuninori.morimoto.gx@renesas.com>
Signed-off-by: Vinod Koul <vinod.koul@intel.com>
The original name was simply hash_string(), but that conflicted with a
function with that name in drivers/base/power/trace.c, and I decided
that calling it "hashlen_" was better anyway.
But you have to do it in two places.
[ This caused build errors for architectures that don't define
CONFIG_DCACHE_WORD_ACCESS - Linus ]
Signed-off-by: George Spelvin <linux@sciencehorizons.net>
Reported-by: Guenter Roeck <linux@roeck-us.net>
Fixes: fcfd2fbf22 ("fs/namei.c: Add hashlen_string() function")
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
The HPFS filesystem used generic_show_options to produce string that is
displayed in /proc/mounts. However, there is a problem that the options
may disappear after remount. If we mount the filesystem with option1
and then remount it with option2, /proc/mounts should show both option1
and option2, however it only shows option2 because the whole option
string is replaced with replace_mount_options in hpfs_remount_fs.
To fix this bug, implement the hpfs_show_options function that prints
options that are currently selected.
Signed-off-by: Mikulas Patocka <mpatocka@redhat.com>
Cc: stable@vger.kernel.org
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Commit c8f33d0bec ("affs: kstrdup() memory handling") checks if the
kstrdup function returns NULL due to out-of-memory condition.
However, if we are remounting a filesystem with no change to
filesystem-specific options, the parameter data is NULL. In this case,
kstrdup returns NULL (because it was passed NULL parameter), although no
out of memory condition exists. The mount syscall then fails with
ENOMEM.
This patch fixes the bug. We fail with ENOMEM only if data is non-NULL.
The patch also changes the call to replace_mount_options - if we didn't
pass any filesystem-specific options, we don't call
replace_mount_options (thus we don't erase existing reported options).
Fixes: c8f33d0bec ("affs: kstrdup() memory handling")
Signed-off-by: Mikulas Patocka <mpatocka@redhat.com>
Cc: stable@vger.kernel.org # v4.1+
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Commit ce657611ba ("hpfs: kstrdup() out of memory handling") checks if
the kstrdup function returns NULL due to out-of-memory condition.
However, if we are remounting a filesystem with no change to
filesystem-specific options, the parameter data is NULL. In this case,
kstrdup returns NULL (because it was passed NULL parameter), although no
out of memory condition exists. The mount syscall then fails with
ENOMEM.
This patch fixes the bug. We fail with ENOMEM only if data is non-NULL.
The patch also changes the call to replace_mount_options - if we didn't
pass any filesystem-specific options, we don't call
replace_mount_options (thus we don't erase existing reported options).
Fixes: ce657611ba ("hpfs: kstrdup() out of memory handling")
Signed-off-by: Mikulas Patocka <mpatocka@redhat.com>
Cc: stable@vger.kernel.org
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Various builds (such as i386:allmodconfig) fail with
fs/binfmt_aout.c:133:2: error: expected identifier or '(' before 'return'
fs/binfmt_aout.c:134:1: error: expected identifier or '(' before '}' token
[ Oops. My bad, I had stupidly thought that "allmodconfig" covered this
on x86-64 too, but it obviously doesn't. Egg on my face. - Linus ]
Fixes: 5d22fc25d4 ("mm: remove more IS_ERR_VALUE abuses")
Signed-off-by: Guenter Roeck <linux@roeck-us.net>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Pull string hash improvements from George Spelvin:
"This series does several related things:
- Makes the dcache hash (fs/namei.c) useful for general kernel use.
(Thanks to Bruce for noticing the zero-length corner case)
- Converts the string hashes in <linux/sunrpc/svcauth.h> to use the
above.
- Avoids 64-bit multiplies in hash_64() on 32-bit platforms. Two
32-bit multiplies will do well enough.
- Rids the world of the bad hash multipliers in hash_32.
This finishes the job started in commit 689de1d6ca ("Minimal
fix-up of bad hashing behavior of hash_64()")
The vast majority of Linux architectures have hardware support for
32x32-bit multiply and so derive no benefit from "simplified"
multipliers.
The few processors that do not (68000, h8/300 and some models of
Microblaze) have arch-specific implementations added. Those
patches are last in the series.
- Overhauls the dcache hash mixing.
The patch in commit 0fed3ac866 ("namei: Improve hash mixing if
CONFIG_DCACHE_WORD_ACCESS") was an off-the-cuff suggestion.
Replaced with a much more careful design that's simultaneously
faster and better. (My own invention, as there was noting suitable
in the literature I could find. Comments welcome!)
- Modify the hash_name() loop to skip the initial HASH_MIX(). This
would let us salt the hash if we ever wanted to.
- Sort out partial_name_hash().
The hash function is declared as using a long state, even though
it's truncated to 32 bits at the end and the extra internal state
contributes nothing to the result. And some callers do odd things:
- fs/hfs/string.c only allocates 32 bits of state
- fs/hfsplus/unicode.c uses it to hash 16-bit unicode symbols not bytes
- Modify bytemask_from_count to handle inputs of 1..sizeof(long)
rather than 0..sizeof(long)-1. This would simplify users other
than full_name_hash"
Special thanks to Bruce Fields for testing and finding bugs in v1. (I
learned some humbling lessons about "obviously correct" code.)
On the arch-specific front, the m68k assembly has been tested in a
standalone test harness, I've been in contact with the Microblaze
maintainers who mostly don't care, as the hardware multiplier is never
omitted in real-world applications, and I haven't heard anything from
the H8/300 world"
* 'hash' of git://ftp.sciencehorizons.net/linux:
h8300: Add <asm/hash.h>
microblaze: Add <asm/hash.h>
m68k: Add <asm/hash.h>
<linux/hash.h>: Add support for architecture-specific functions
fs/namei.c: Improve dcache hash function
Eliminate bad hash multipliers from hash_32() and hash_64()
Change hash_64() return value to 32 bits
<linux/sunrpc/svcauth.h>: Define hash_str() in terms of hashlen_string()
fs/namei.c: Add hashlen_string() function
Pull out string hash to <linux/stringhash.h>
This will improve the performance of hash_32() and hash_64(), but due
to complete lack of multi-bit shift instructions on H8, performance will
still be bad in surrounding code.
Designing H8-specific hash algorithms to work around that is a separate
project. (But if the maintainers would like to get in touch...)
Signed-off-by: George Spelvin <linux@sciencehorizons.net>
Cc: Yoshinori Sato <ysato@users.sourceforge.jp>
Cc: uclinux-h8-devel@lists.sourceforge.jp
