An error occurred fetching the project authors.
- 31 Jan, 2011 13 commits
-
-
Rabin Vincent authored
To simplify the code. Acked-by:
Per Forlin <per.forlin@stericsson.com> Acked-by:
Jonas Aaberg <jonas.aberg@stericsson.com> Signed-off-by:
Rabin Vincent <rabin.vincent@stericsson.com> Signed-off-by:
Linus Walleij <linus.walleij@stericsson.com> Signed-off-by:
Dan Williams <dan.j.williams@intel.com>
-
Rabin Vincent authored
Acked-by:
Per Forlin <per.forlin@stericsson.com> Acked-by:
Jonas Aaberg <jonas.aberg@stericsson.com> Signed-off-by:
Rabin Vincent <rabin.vincent@stericsson.com> Signed-off-by:
Linus Walleij <linus.walleij@stericsson.com> Signed-off-by:
Dan Williams <dan.j.williams@intel.com>
-
Rabin Vincent authored
Map and unmap the LLIs and use dma_sync_single_for_device() appropriately instead of mapping and never unmapping them. Acked-by:
Per Forlin <per.forlin@stericsson.com> Acked-by:
Jonas Aaberg <jonas.aberg@stericsson.com> Signed-off-by:
Rabin Vincent <rabin.vincent@stericsson.com> Signed-off-by:
Linus Walleij <linus.walleij@stericsson.com> Signed-off-by:
Dan Williams <dan.j.williams@intel.com>
-
Rabin Vincent authored
Map the buffer once and use dma_sync*() appropriately instead of mapping the buffer over and over without unmapping it. Acked-by:
Per Forlin <per.forlin@stericsson.com> Acked-by:
Jonas Aaberg <jonas.aberg@stericsson.com> Signed-off-by:
Rabin Vincent <rabin.vincent@stericsson.com> Signed-off-by:
Linus Walleij <linus.walleij@stericsson.com> Signed-off-by:
Dan Williams <dan.j.williams@intel.com>
-
Rabin Vincent authored
ALIGN(x * y, y) == x * y ALIGN(aligned + x * y, y) == aligned + x * y Acked-by:
Per Forlin <per.forlin@stericsson.com> Acked-by:
Jonas Aaberg <jonas.aberg@stericsson.com> Signed-off-by:
Rabin Vincent <rabin.vincent@stericsson.com> Signed-off-by:
Linus Walleij <linus.walleij@stericsson.com> Signed-off-by:
Dan Williams <dan.j.williams@intel.com>
-
Rabin Vincent authored
DB8500v2's DMA40 (revision 3) allows setting event lines as high priority and real time. Acked-by:
Per Forlin <per.forlin@stericsson.com> Acked-by:
Jonas Aaberg <jonas.aberg@stericsson.com> Signed-off-by:
Rabin Vincent <rabin.vincent@stericsson.com> Signed-off-by:
Linus Walleij <linus.walleij@stericsson.com> Signed-off-by:
Dan Williams <dan.j.williams@intel.com>
-
Rabin Vincent authored
And add DB8500v2 information. Acked-by:
Per Forlin <per.forlin@stericsson.com> Acked-by:
Jonas Aaberg <jonas.aberg@stericsson.com> Signed-off-by:
Rabin Vincent <rabin.vincent@stericsson.com> Signed-off-by:
Linus Walleij <linus.walleij@stericsson.com> Signed-off-by:
Dan Williams <dan.j.williams@intel.com>
-
Rabin Vincent authored
Almost every use of dev_err in this driver prints the function name. Abstract out wrappers to help with this and reduce code duplication. Acked-by:
Per Forlin <per.forlin@stericsson.com> Acked-by:
Jonas Aaberg <jonas.aberg@stericsson.com> Signed-off-by:
Rabin Vincent <rabin.vincent@stericsson.com> Signed-off-by:
Linus Walleij <linus.walleij@stericsson.com> Signed-off-by:
Dan Williams <dan.j.williams@intel.com>
-
Rabin Vincent authored
The somewhat confusing check d40c->log_num == D40_PHY_CHAN and its variants are used in several places to check if a channel is logical or physical. Use appropriately named helpers to do this to make the code more readable. Acked-by:
Per Forlin <per.forlin@stericsson.com> Acked-by:
Jonas Aaberg <jonas.aberg@stericsson.com> Signed-off-by:
Rabin Vincent <rabin.vincent@stericsson.com> Signed-off-by:
Linus Walleij <linus.walleij@stericsson.com> Signed-off-by:
Dan Williams <dan.j.williams@intel.com>
-
Rabin Vincent authored
The register offset computation for accessing channel registers is copy/pasted in several places. Create a helper function to do it. Acked-by:
Per Forlin <per.forlin@stericsson.com> Acked-by:
Jonas Aaberg <jonas.aberg@stericsson.com> Signed-off-by:
Rabin Vincent <rabin.vincent@stericsson.com> Signed-off-by:
Linus Walleij <linus.walleij@stericsson.com> Signed-off-by:
Dan Williams <dan.j.williams@intel.com>
-
Rabin Vincent authored
This link in hardware with previous jobs code is: - unused, no clients using or requiring this feature - incomplete, being implemented only for physical channels - broken, only working to perform one link Remove it. This also allows us to get rid of the channel pause in the submit_tx() routine. Acked-by:
Per Forlin <per.forlin@stericsson.com> Acked-by:
Jonas Aaberg <jonas.aberg@stericsson.com> Signed-off-by:
Rabin Vincent <rabin.vincent@stericsson.com> Signed-off-by:
Linus Walleij <linus.walleij@stericsson.com> Signed-off-by:
Dan Williams <dan.j.williams@intel.com>
-
Rabin Vincent authored
The controller sometimes fails to register the enable of the event line when both src and dst event lines are used on the same logical channel. Implement the recommended software workaround, which is to retry the write until it works. Acked-by:
Per Forlin <per.forlin@stericsson.com> Acked-by:
Jonas Aaberg <jonas.aberg@stericsson.com> Signed-off-by:
Rabin Vincent <rabin.vincent@stericsson.com> Signed-off-by:
Linus Walleij <linus.walleij@stericsson.com> Signed-off-by:
Dan Williams <dan.j.williams@intel.com>
-
Rabin Vincent authored
Acked-by:
Per Forlin <per.forlin@stericsson.com> Acked-by:
Jonas Aaberg <jonas.aberg@stericsson.com> Signed-off-by:
Rabin Vincent <rabin.vincent@stericsson.com> Signed-off-by:
Linus Walleij <linus.walleij@stericsson.com> Signed-off-by:
Dan Williams <dan.j.williams@intel.com>
-
- 05 Jan, 2011 1 commit
-
-
Per Forlin authored
The maximum transfer size of the stedma40 is (64k-1) x data-width. If the transfer size of one element exceeds this limit the job is split up and sent as linked transfer. Signed-off-by:
Per Forlin <per.forlin@linaro.org> Signed-off-by:
Dan Williams <dan.j.williams@intel.com>
-
- 19 Oct, 2010 4 commits
-
-
Rabin Vincent authored
A bool will suffice. The default is little endian. Acked-by:
Jonas Aaberg <jonas.aberg@stericsson.com> Signed-off-by:
Rabin Vincent <rabin.vincent@stericsson.com> Signed-off-by:
Linus Walleij <linus.walleij@stericsson.com> Signed-off-by:
Dan Williams <dan.j.williams@intel.com>
-
Rabin Vincent authored
Defaults are "basic mode" for physical channels, and "logical source logical destination" for logical channels. Acked-by:
Jonas Aaberg <jonas.aberg@stericsson.com> Signed-off-by:
Rabin Vincent <rabin.vincent@stericsson.com> Signed-off-by:
Linus Walleij <linus.walleij@stericsson.com> Signed-off-by:
Dan Williams <dan.j.williams@intel.com>
-
Rabin Vincent authored
And keep it logical by default. Acked-by:
Jonas Aaberg <jonas.aberg@stericsson.com> Signed-off-by:
Rabin Vincent <rabin.vincent@stericsson.com> Signed-off-by:
Linus Walleij <linus.walleij@stericsson.com> Signed-off-by:
Dan Williams <dan.j.williams@intel.com>
-
Rabin Vincent authored
Since we want to reduce the amount of required channel configuration and remove channel_type, don't depend on it to indicate whether the configuration is valid. Acked-by:
Jonas Aaberg <jonas.aberg@stericsson.com> Signed-off-by:
Rabin Vincent <rabin.vincent@stericsson.com> Signed-off-by:
Linus Walleij <linus.walleij@stericsson.com> Signed-off-by:
Dan Williams <dan.j.williams@intel.com>
-
- 07 Oct, 2010 7 commits
-
-
Per Forlin authored
Remove obsolete pre_transfer hook in stedma40_chan_cfg. The intent of this hook is merely to handle burst size compensation for ux500 variant MMCI. Remove obsolete stedma40_set_psize since it is only called from pre_transfer. DMAEngine device_control replaces the functionality of stedma40_set_psize. Signed-off-by:
Per Forlin <per.forlin@stericsson.com> Signed-off-by:
Dan Williams <dan.j.williams@intel.com>
-
Per Forlin authored
Set burst for physical or logical channels respectively. Convert the values in dma_cfg to dma reg bits for physical or logical channels. Signed-off-by:
Per Forlin <per.forlin@stericsson.com> Signed-off-by:
Linus Walleij <linus.walleij@stericsson.com> Signed-off-by:
Dan Williams <dan.j.williams@intel.com>
-
Rabin Vincent authored
Fix some leaks of allocated descriptors in error paths. Acked-by:
Jonas Aaberg <jonas.aberg@stericsson.com> Signed-off-by:
Rabin Vincent <rabin.vincent@stericsson.com> Signed-off-by:
Linus Walleij <linus.walleij@stericsson.com> Signed-off-by:
Dan Williams <dan.j.williams@intel.com>
-
Rabin Vincent authored
Fix desc_get to alloc a descriptor from the cache if the ones in the list are waiting for the ack. Also, memzero the descriptor when allocated from the list to ensure all fields are cleared. Acked-by:
Jonas Aaberg <jonas.aberg@stericsson.com> Signed-off-by:
Rabin Vincent <rabin.vincent@stericsson.com> Signed-off-by:
Linus Walleij <linus.walleij@stericsson.com> Signed-off-by:
Dan Williams <dan.j.williams@intel.com>
-
Rabin Vincent authored
clk_get returns an ERR_PTR. Acked-by:
Jonas Aaberg <jonas.aberg@stericsson.com> Signed-off-by:
Rabin Vincent <rabin.vincent@stericsson.com> Signed-off-by:
Linus Walleij <linus.walleij@stericsson.com> Signed-off-by:
Dan Williams <dan.j.williams@intel.com>
-
Rabin Vincent authored
The value in the array, not the index, specifies the channel to be disabled. Acked-by:
Jonas Aaberg <jonas.aberg@stericsson.com> Signed-off-by:
Rabin Vincent <rabin.vincent@stericsson.com> Signed-off-by:
Linus Walleij <linus.walleij@stericsson.com> Signed-off-by:
Dan Williams <dan.j.williams@intel.com>
-
Ira Snyder authored
Now that the DMAEngine API has support for scatterlist to scatterlist copy, implement support for the STE DMA40 DMA controller. Cc: Linus Walleij <linus.ml.walleij@gmail.com> Acked-by:
Per Fridén <per.friden@stericsson.com> Signed-off-by:
Ira W. Snyder <iws@ovro.caltech.edu> Signed-off-by:
Dan Williams <dan.j.williams@intel.com>
-
- 22 Sep, 2010 12 commits
-
-
Jonas Aaberg authored
There is no point in pausing what isn't running. Signed-off-by:
Jonas Aaberg <jonas.aberg@stericsson.com> Signed-off-by:
Linus Walleij <linus.walleij@stericsson.com> Signed-off-by:
Dan Williams <dan.j.williams@intel.com>
-
Jonas Aaberg authored
LLI allocation is now done on job level instead of channel level. Previously the maximum length of a linked job in hw on a logical channel was 8, since the LLIs where evenly divided. Now only executing jobs have allocated LLIs which increase the length to a maximum of 64 links in HW. Signed-off-by:
Jonas Aaberg <jonas.aberg@stericsson.com> Signed-off-by:
Linus Walleij <linus.walleij@stericsson.com> Signed-off-by:
Dan Williams <dan.j.williams@intel.com>
-
Jonas Aaberg authored
Signed-off-by:
Jonas Aaberg <jonas.aberg@stericsson.com> Signed-off-by:
Linus Walleij <linus.walleij@stericsson.com> Signed-off-by:
Dan Williams <dan.j.williams@intel.com>
-
Jonas Aaberg authored
The handling of pause detection was slightly incorrect. Signed-off-by:
Jonas Aaberg <jonas.aberg@stericsson.com> Signed-off-by:
Linus Walleij <linus.walleij@stericsson.com> Signed-off-by:
Dan Williams <dan.j.williams@intel.com>
-
Jonas Aaberg authored
This patch includes non functional code clean up changes, file header updates and a few magic numbers got defined. Signed-off-by:
Jonas Aaberg <jonas.aberg@stericsson.com> Signed-off-by:
Linus Walleij <linus.walleij@stericsson.com> Signed-off-by:
Dan Williams <dan.j.williams@intel.com>
-
Jonas Aaberg authored
If a new job is added on a physical channel that already has a job, the new job is linked in hw to the old job instead of queueing up the jobs. Signed-off-by:
Jonas Aaberg <jonas.aberg@stericsson.com> Signed-off-by:
Linus Walleij <linus.walleij@stericsson.com> Signed-off-by:
Dan Williams <dan.j.williams@intel.com>
-
Jonas Aaberg authored
Make sure to extract the revision field explicitly and document what bits are being accessed here without magic numbers. Signed-off-by:
Jonas Aaberg <jonas.aberg@stericsson.com> Signed-off-by:
Linus Walleij <linus.walleij@stericsson.com> Signed-off-by:
Dan Williams <dan.j.williams@intel.com>
-
Jonas Aaberg authored
The callback got called even when it was not supposed to. Also removed some not needed interrupt trigger on/off code. Signed-off-by:
Jonas Aaberg <jonas.aberg@stericsson.com> Signed-off-by:
Linus Walleij <linus.walleij@stericsson.com> Signed-off-by:
Dan Williams <dan.j.williams@intel.com>
-
Jonas Aaberg authored
This patch cleans up some code and removes a suspend request that was pointless since the hw was never configured nor running when it was called. Signed-off-by:
Jonas Aaberg <jonas.aberg@stericsson.com> Signed-off-by:
Linus Walleij <linus.walleij@stericsson.com> Signed-off-by:
Dan Williams <dan.j.williams@intel.com>
-
Jonas Aaberg authored
Removes an unnecessary register read and a few lines of code. Signed-off-by:
Jonas Aaberg <jonas.aberg@stericsson.com> Signed-off-by:
Linus Walleij <linus.walleij@stericsson.com> Signed-off-by:
Dan Williams <dan.j.williams@intel.com>
-
Jonas Aaberg authored
A transfer that runs in the different direction on the same channel will now be resumed when the other is suspend/stopped. Signed-off-by:
Jonas Aaberg <jonas.aberg@stericsson.com> Signed-off-by:
Linus Walleij <linus.walleij@stericsson.com> Signed-off-by:
Dan Williams <dan.j.williams@intel.com>
-
Linus Walleij authored
Added various configuration checks. Signed-off-by:
Jonas Aaberg <jonas.aberg@stericsson.com> Signed-off-by:
Linus Walleij <linus.walleij@stericsson.com> Signed-off-by:
Dan Williams <dan.j.williams@intel.com>
-
- 04 Aug, 2010 1 commit
-
-
Linus Walleij authored
This extends the DMA engine driver for the DMA40 used in the U8500 platform with the generic runtime slave configuration interface. Signed-off-by:
Linus Walleij <linus.walleij@stericsson.com> Signed-off-by:
Dan Williams <dan.j.williams@intel.com>
-
- 23 Jun, 2010 2 commits
-
-
Linus Walleij authored
This makes sure the DMA40 driver will also work on the oldest silicon revisions that have the on-chip memory on another location in the DB8500 and also requires explicit suspend before starting or resuming a logical channel. Signed-off-by:
Linus Walleij <linus.walleij@stericsson.com> [added parenthesis to the definition of U8500_DMA_LCPA_BASE_ED] Signed-off-by:
Dan Williams <dan.j.williams@intel.com>
-
Jonas Aaberg authored
This makes it possible to disable physical channels. Signed-off-by:
Jonas Aaberg <jonas.aberg@stericsson.com> Signed-off-by:
Linus Walleij <linus.walleij@stericsson.com> Signed-off-by:
Dan Williams <dan.j.williams@intel.com>
-