linux-toradex.git/drivers/dma, branch v4.2.7

dmaengine: pxa_dma: fix initial list move

2015-10-22T21:49:17+00:00

commit aebf5a67db8dbacbc624b9c652b81f5460b15eff upstream.

Since the commit to have an allocated list of virtual descriptors was
reverted, the pxa_dma driver is broken, as it assumes the descriptor is
placed on the allocated list upon allocation.

Fix the issue in pxa_dma by making an allocated virtual descriptor a
singleton.

Fixes: 8c8fe97b2b8a ("Revert "dmaengine: virt-dma: don't always free descriptor upon completion"")
Signed-off-by: Robert Jarzmik 
Signed-off-by: Vinod Koul 
Signed-off-by: Greg Kroah-Hartman

dmaengine: at_xdmac: fix bug in prep_dma_cyclic

2015-10-22T21:49:17+00:00

commit e900c30dc1bb0cbc07708e9be1188f531632b2ef upstream.

In cyclic mode, the round chaining has been broken by the introduction
of at_xdmac_queue_desc(): AT_XDMAC_MBR_UBC_NDE is set for all descriptors
excepted for the last one. at_xdmac_queue_desc() has to be called one
more time to chain the last and the first descriptors.

Signed-off-by: Ludovic Desroches 
Fixes: 0d0ee751f7f7 ("dmaengine: xdmac: Rework the chaining logic")
Signed-off-by: Vinod Koul 
Signed-off-by: Greg Kroah-Hartman

dmaengine: dw: properly read DWC_PARAMS register

2015-10-22T21:49:17+00:00

commit 6bea0f6d1c47b07be88dfd93f013ae05fcb3d8bf upstream.

In case we have less than maximum allowed channels (8) and autoconfiguration is
enabled the DWC_PARAMS read is wrong because it uses different arithmetic to
what is needed for channel priority setup.

Re-do the caclulations properly. This now works on AVR32 board well.

Fixes: fed2574b3c9f (dw_dmac: introduce software emulation of LLP transfers)
Cc: yitian.bu@tangramtek.com
Signed-off-by: Andy Shevchenko 
Signed-off-by: Vinod Koul 
Signed-off-by: Greg Kroah-Hartman

dmaengine: at_xdmac: clean used descriptor

2015-10-22T21:49:17+00:00

commit 0be2136b67067617b36c70e525d7534108361e36 upstream.

When putting back a descriptor to the free descs list, some fields are
not set to 0, it can cause bugs if someone uses it without having this
in mind.
Descriptor are not put back one by one so it is easier to clean
descriptors when we request them.

Signed-off-by: Ludovic Desroches 
Signed-off-by: Vinod Koul 
Signed-off-by: Greg Kroah-Hartman

dmaengine: at_xdmac: change block increment addressing mode

2015-10-22T21:49:17+00:00

commit a1cf09031e641d3cceaca4a4dd20ef6a785bc9b3 upstream.

The addressing mode we were using was not only incrementing the address at
each microblock, but also at each data boundary, which was severely slowing
the transfer, without any benefit since we were not using the data stride.

Switch to the micro block increment only in order to get back to an
acceptable performance level.

Signed-off-by: Maxime Ripard 
Signed-off-by: Ludovic Desroches 
Fixes: 6007ccb57744 ("dmaengine: xdmac: Add interleaved transfer support")
Signed-off-by: Vinod Koul 
Signed-off-by: Greg Kroah-Hartman

dmaengine: fix balance of privatecnt inc/dec operations

2015-08-17T17:17:43+00:00

This patch increments privatecnt value and set DMA_PRIVATE in device
caps in dma_request_slave_channel() function. This is needed to keep
privatecnt increment/decrement balance.

As function dma_release_channel() decrements privatecnt counter, we need
to increment it when channel is requested. Otherwise privatecnt drops
into negatives after few dma_release_channel() calls.

Reported-by: Krzysztof Kozlowski 
Signed-off-by: Robert Baldyga 
Signed-off-by: Vinod Koul

Merge tag 'dmaengine-fix-4.2-rc5' of git://git.infradead.org/users/vkoul/slave-dma

2015-08-01T19:47:04+00:00

Pull dmaengine fixes from Vinod Koul:
 "We had a regression due to reuse of descriptor so we have reverted
  that.

  The rest are driver fixes:

   - at_hdmac and at_xdmac for residue, trannfer width, and channel config
   - pl330 final fix for dma fails and overflow issue
   - xgene resouce map fix
   - mv_xor big endian op fix"

* tag 'dmaengine-fix-4.2-rc5' of git://git.infradead.org/users/vkoul/slave-dma:
  Revert "dmaengine: virt-dma: don't always free descriptor upon completion"
  dmaengine: mv_xor: fix big endian operation in register mode
  dmaengine: xgene-dma: Fix the resource map to handle overlapping
  dmaengine: at_xdmac: fix transfer data width in at_xdmac_prep_slave_sg()
  dmaengine: at_hdmac: fix residue computation
  dmaengine: at_xdmac: fix bug about channel configuration
  dmaengine: pl330: Really fix choppy sound because of wrong residue calculation
  dmaengine: pl330: Fix overflow when reporting residue in memcpy

Revert "dmaengine: virt-dma: don't always free descriptor upon completion"

2015-07-31T15:03:43+00:00

This reverts commit b9855f03d560d351e95301b9de0bc3cad3b31fe9.
The patch break existing DMA usage case. For example, audio SOC
dmaengine never release channel and cause virt-dma to cache too
much memory in descriptor to exhaust system memory.

Signed-off-by: Vinod Koul

dmaengine: mv_xor: fix big endian operation in register mode

2015-07-31T15:03:43+00:00

Commit 6f166312c6ea2 ("dmaengine: mv_xor: add support for a38x command
in descriptor mode") introduced the support for a feature that
appeared in Armada 38x: specifying the operation to be performed in a
per-descriptor basis rather than globally per channel.

However, when doing so, it changed the function mv_chan_set_mode() to
use:

  if (IS_ENABLED(__BIG_ENDIAN))

instead of:

  #if defined(__BIG_ENDIAN)

While IS_ENABLED() is perfectly fine for CONFIG_* symbols, it is not
for other symbols such as __BIG_ENDIAN that is provided directly by
the compiler. Consequently, the commit broke support for big-endian,
as the XOR_DESCRIPTOR_SWAP flag was not set in the XOR channel
configuration register.

The primarily visible effect was some nasty warnings and failures
appearing during the self-test of the XOR unit:

[    1.197368] mv_xor d0060900.xor: error on chan 0. intr cause 0x00000082
[    1.197393] mv_xor d0060900.xor: config       0x00008440
[    1.197410] mv_xor d0060900.xor: activation   0x00000000
[    1.197427] mv_xor d0060900.xor: intr cause   0x00000082
[    1.197443] mv_xor d0060900.xor: intr mask    0x000003f7
[    1.197460] mv_xor d0060900.xor: error cause  0x00000000
[    1.197477] mv_xor d0060900.xor: error addr   0x00000000
[    1.197491] ------------[ cut here ]------------
[    1.197513] WARNING: CPU: 0 PID: 1 at ../drivers/dma/mv_xor.c:664 mv_xor_interrupt_handler+0x14c/0x170()

See also:

  http://storage.kernelci.org/next/next-20150617/arm-mvebu_v7_defconfig+CONFIG_CPU_BIG_ENDIAN=y/lab-khilman/boot-armada-xp-openblocks-ax3-4.txt

Signed-off-by: Thomas Petazzoni 
Fixes: 6f166312c6ea2 ("dmaengine: mv_xor: add support for a38x command in descriptor mode")
Reviewed-by: Maxime Ripard 
Signed-off-by: Vinod Koul

dmaengine: xgene-dma: Fix the resource map to handle overlapping

2015-07-31T15:03:43+00:00

There is an overlap in dma ring cmd csr region due to sharing of ethernet
ring cmd csr region. This patch fix the resource overlapping by mapping
the entire dma ring cmd csr region.

Signed-off-by: Rameshwar Prasad Sahu 
Signed-off-by: Vinod Koul