Commit graph

81 commits

Author SHA1 Message Date
Dinh Nguyen
47910a49db fpga: fix sparse warnings in fpga-mgr and fpga-bridge
Fix up these sparse warnings:

drivers/fpga/fpga-mgr.c:189:21: warning: symbol '__fpga_mgr_get' was not declared. Should it be static?
drivers/fpga/fpga-bridge.c:30:12: warning: symbol 'bridge_list_lock' was not declared. Should it be static?

Signed-off-by: Dinh Nguyen <dinguyen@kernel.org>
Acked-by: Moritz Fischer <mdf@kernel.org>
Signed-off-by: Alan Tull <atull@kernel.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
2017-03-17 15:10:48 +09:00
Jason Gunthorpe
425902f5c8 fpga zynq: Use the scatterlist interface
This allows the driver to avoid a high order coherent DMA allocation
and memory copy. With this patch it can DMA directly from the kernel
pages that the bitfile is stored in.

Since this is now a gather DMA operation the driver uses the ISR
to feed the chips DMA queue with each entry from the SGL.

Signed-off-by: Jason Gunthorpe <jgunthorpe@obsidianresearch.com>
Acked-by: Moritz Fischer <moritz.fischer@ettus.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
2017-02-10 15:20:44 +01:00
Jason Gunthorpe
baa6d39663 fpga: Add scatterlist based programming
Requiring contiguous kernel memory is not a good idea, this is a limited
resource and allocation can fail under normal work loads.

This introduces a .write_sg op that supporting drivers can provide
to DMA directly from dis-contiguous memory and a new entry point
fpga_mgr_buf_load_sg that users can call to directly provide page
lists.

The full matrix of compatibility is provided, either the linear or sg
interface can be used by the user with a driver supporting either
interface.

A notable change for drivers is that the .write op can now be called
multiple times.

Signed-off-by: Jason Gunthorpe <jgunthorpe@obsidianresearch.com>
Acked-by: Alan Tull <atull@opensource.altera.com>
Acked-by: Moritz Fischer <moritz.fischer@ettus.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
2017-02-10 15:20:44 +01:00
Jason Gunthorpe
b496df86ac fpga zynq: Check the bitstream for validity
There is no sense in sending a bitstream we know will not work, and
with the variety of options for bitstream generation in Xilinx tools
it is not terribly clear what the correct input should be.

This is particularly important for Zynq since auto-correction was
removed from the driver and the Zynq hardware only accepts a bitstream
format that is different from what the Xilinx tools typically produce.

Worse, the hardware provides no indication why the bitstream fails,
it simply times out if the input is wrong.

The best option here is to have the kernel print a message informing
the user they are using a malformed bistream and programming failure
isn't for any of the myriad of other reasons.

Signed-off-by: Jason Gunthorpe <jgunthorpe@obsidianresearch.com>
Acked-by: Moritz Fischer <moritz.fischer@ettus.com>
Acked-by: Alan Tull <atull@opensource.altera.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
2017-02-10 15:20:44 +01:00
Jason Gunthorpe
6b45e0f24c fpga zynq: Check for errors after completing DMA
The completion did not check the interrupt status to see if any error
bits were asserted, check error bits and dump some registers if things
went wrong.

A few fixes are needed to make this work, the IXR_ERROR_FLAGS_MASK was
wrong, it included the done bits, which shows a bug in mask/unmask_irqs
which were using the wrong bits, simplify all of this stuff.

Signed-off-by: Jason Gunthorpe <jgunthorpe@obsidianresearch.com>
Reviewed-by: Moritz Fischer <moritz.fischer@ettus.com>
Acked-by: Alan Tull <atull@opensource.altera.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
2017-02-10 15:20:44 +01:00
Jason Gunthorpe
1d7f1589d3 fpga: Clarify how write_init works streaming modes
This interface was designed for streaming, but write_init's buf
argument has an unclear purpose. Define it to be the first bytes
of the bitstream. Each driver gets to set how many bytes (at most)
it wants to see. Short bitstreams will be passed through as-is, while
long ones will be truncated.

The intent is to allow drivers to peek at the header before the transfer
actually starts.

Signed-off-by: Jason Gunthorpe <jgunthorpe@obsidianresearch.com>
Acked-by: Alan Tull <atull@opensource.altera.com>
2016-11-29 15:51:49 -06:00
Jason Gunthorpe
340c0c53ea fpga zynq: Fix incorrect ISR state on bootup
It is best practice to clear and mask all interrupts before
associating the IRQ, and this should be done after the clock
is enabled.

This corrects a bad result from zynq_fpga_ops_state on bootup
where left over latched values in INT_STS_OFFSET caused it to
report an unconfigured FPGA as configured.

After this change the boot up operating state for an unconfigured
FPGA reports 'unknown'.

Signed-off-by: Jason Gunthorpe <jgunthorpe@obsidianresearch.com>
Acked-by: Alan Tull <atull@opensource.altera.com>
Acked-by: Moritz Fischer <moritz.fischer@ettus.com>
2016-11-29 15:51:48 -06:00
Jason Gunthorpe
80baf649c2 fpga zynq: Remove priv->dev
socfpga uses mgr->dev for debug prints, there should be consistency
here, so standardize on that. The only other use was for dma
which can be replaced with mgr->dev.parent.

Signed-off-by: Jason Gunthorpe <jgunthorpe@obsidianresearch.com>
Acked-by: Alan Tull <atull@opensource.altera.com>
Acked-by: Moritz Fischer <moritz.fischer@ettus.com>
2016-11-29 15:51:46 -06:00
Jason Gunthorpe
1930c28651 fpga zynq: Add missing \n to messages
Function dev_err doesn't add a newline at the end of the string. This will
lead to a hard to read kernel log.

Signed-off-by: Jason Gunthorpe <jgunthorpe@obsidianresearch.com>
Reviewed-by: Moritz Fischer <moritz.fischer@ettus.com>
Reviewed-by: Matthias Brugger <mbrugger@suse.com>
Acked-by: Alan Tull <atull@opensource.altera.com>
2016-11-29 15:51:45 -06:00
Jason Gunthorpe
a0e1b61858 fpga: Add COMPILE_TEST to all drivers
Like Zynq the Altera drivers compile fine on x86 and others too,
so make it easier to compile test this stuff.

A10 requires REGMAP_MMIO to compile, so be explicit rather than
relying on it via ARCH_SOCFPGA.

Signed-off-by: Jason Gunthorpe <jgunthorpe@obsidianresearch.com>
Acked-by: Alan Tull <atull@opensource.altera.com>
2016-11-29 15:51:44 -06:00
Dan Carpenter
96bc818397 ARM: socfpga: checking the wrong variable
This is a cut and paste bug.  We had intended to check "sysmgr".

Fixes: e5f8efa5c8 ("ARM: socfpga: fpga bridge driver support")
Signed-off-by: Dan Carpenter <dan.carpenter@oracle.com>
Acked-by: Moritz Fischer <moritz.fischer@ettus.com>
Acked-by: Alan Tull <atull@opensource.altera.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
2016-11-17 08:14:55 +01:00
Alan Tull
acbb910ae0 fpga-manager: Add Socfpga Arria10 support
Add low level driver to support reprogramming FPGAs for Altera
SoCFPGA Arria10.

Signed-off-by: Alan Tull <atull@opensource.altera.com>
Reviewed-by: Moritz Fischer <moritz.fischer@ettus.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
2016-11-10 17:03:36 +01:00
Alan Tull
ca24a648f5 fpga: add altera freeze bridge support
Add a low level driver for Altera Freeze Bridges to the FPGA Bridge
framework.  A freeze bridge is a bridge that exists in the FPGA
fabric to isolate one region of the FPGA from the busses while that
one region is being reprogrammed.

Signed-off-by: Alan Tull <atull@opensource.altera.com>
Signed-off-by: Matthew Gerlach <mgerlach@opensource.altera.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
2016-11-10 17:03:36 +01:00
Alan Tull
e5f8efa5c8 ARM: socfpga: fpga bridge driver support
Supports Altera SOCFPGA bridges:
 * fpga2sdram
 * fpga2hps
 * hps2fpga
 * lwhps2fpga

Allows enabling/disabling the bridges through the FPGA
Bridge Framework API functions.

The fpga2sdram driver only supports enabling and disabling
of the ports that been configured early on.  This is due to
a hardware limitation where the read, write, and command
ports on the fpga2sdram bridge can only be reconfigured
while there are no transactions to the sdram, i.e. when
running out of OCRAM before the kernel boots.

Device tree property 'init-val' configures the driver to
enable or disable the bridge during probe.  If the property
does not exist, the driver will leave the bridge in its
current state.

Signed-off-by: Alan Tull <atull@opensource.altera.com>
Signed-off-by: Matthew Gerlach <mgerlach@altera.com>
Signed-off-by: Dinh Nguyen <dinguyen@opensource.altera.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
2016-11-10 17:03:36 +01:00
Alan Tull
0fa20cdfcc fpga: fpga-region: device tree control for FPGA
FPGA Regions support programming FPGA under control of the Device
Tree.

Signed-off-by: Alan Tull <atull@opensource.altera.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
2016-11-10 17:03:35 +01:00
Alan Tull
21aeda950c fpga: add fpga bridge framework
This framework adds API functions for enabling/
disabling FPGA bridges under kernel control.

This allows the Linux kernel to disable FPGA bridges
during FPGA reprogramming and to enable FPGA bridges
when FPGA reprogramming is done.  This framework is
be manufacturer-agnostic, allowing it to be used in
interfaces that use the FPGA Manager Framework to
reprogram FPGA's.

The functions are:
* of_fpga_bridge_get
* fpga_bridge_put
   Get/put an exclusive reference to a FPGA bridge.

* fpga_bridge_enable
* fpga_bridge_disable
   Enable/Disable traffic through a bridge.

* fpga_bridge_register
* fpga_bridge_unregister
   Register/unregister a device-specific low level FPGA
   Bridge driver.

Get an exclusive reference to a bridge and add it to a list:
* fpga_bridge_get_to_list

To enable/disable/put a set of bridges that are on a list:
* fpga_bridges_enable
* fpga_bridges_disable
* fpga_bridges_put

Signed-off-by: Alan Tull <atull@opensource.altera.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
2016-11-10 17:03:35 +01:00
Alan Tull
1df2865f8d fpga-mgr: add fpga image information struct
This patch adds a minor change in the FPGA Manager API
to hold information that is specific to an FPGA image
file.  This change is expected to bring little, if any,
pain.  The socfpga and zynq drivers are fixed up in
this patch.

An FPGA image file will have particulars that affect how the
image is programmed to the FPGA.  One example is that
current 'flags' currently has one bit which shows whether the
FPGA image was built for full reconfiguration or partial
reconfiguration.  Another example is timeout values for
enabling or disabling the bridges in the FPGA.  As the
complexity of the FPGA design increases, the bridges in the
FPGA may take longer times to enable or disable.

This patch adds a new 'struct fpga_image_info', moves the
current 'u32 flags' to it.  Two other image-specific u32's
are added for the bridge enable/disable timeouts.  The FPGA
Manager API functions are changed, replacing the 'u32 flag'
parameter with a pointer to struct fpga_image_info.
Subsequent patches fix the existing low level FPGA manager
drivers.

Signed-off-by: Alan Tull <atull@opensource.altera.com>
Acked-by: Moritz Fischer <moritz.fischer@ettus.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
2016-11-10 17:03:35 +01:00
Alan Tull
9dce0287a6 fpga: add method to get fpga manager from device
The intent is to provide a non-DT method of getting
ahold of a FPGA manager to do some FPGA programming.

This patch refactors of_fpga_mgr_get() to reuse most of it
while adding a new method fpga_mgr_get() for getting a
pointer to a fpga manager struct, given the device.

Signed-off-by: Alan Tull <atull@opensource.altera.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
2016-11-10 17:03:35 +01:00
Jean Delvare
54e9b09972 fpga manager: Add hardware dependency to Zynq driver
The Zynq FPGA manager driver serves no purpose on other architectures
so hide it unless build-testing.

Signed-off-by: Jean Delvare <jdelvare@suse.de>
Acked-by: Moritz Fischer <moritz.fischer@ettus.com>
Acked-by: Alan Tull <atull@opensource.altera.com>
Acked-by: Michal Simek <michal.simek@xilinx.com>
Cc: "Sören Brinkmann" <soren.brinkmann@xilinx.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
2016-09-08 17:31:44 +02:00
Sudip Mukherjee
1c8cb40949 drivers/fpga/Kconfig: fix build failure
While building m32r allmodconfig the build is failing with the error:

  ERROR: "bad_dma_ops" [drivers/fpga/zynq-fpga.ko] undefined!

Xilinx Zynq FPGA is using DMA but there was no dependency while
building.

Link: http://lkml.kernel.org/r/1464346526-13913-1-git-send-email-sudipm.mukherjee@gmail.com
Signed-off-by: Sudip Mukherjee <sudip.mukherjee@codethink.co.uk>
Acked-by: Moritz Fischer <moritz.fischer@ettus.com>
Cc: Alan Tull <atull@opensource.altera.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2016-08-04 08:50:07 -04:00
Tobias Klauser
e8c77bda05 fpga manager: Fix firmware resource leak on error
If fpga_mgr_buf_load() fails, the firmware resource previously allocated
by request_firmware() is leaked. Fix it by calling release_firmware()
regardless of the return value of fpga_mgr_buf_load().

Found by the Coverity scanner (CID 1339653).

Fixes: 6a8c3be7ec ("add FPGA manager core")
Signed-off-by: Tobias Klauser <tklauser@distanz.ch>
Acked-by: Alan Tull <atull@opensource.altera.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
2015-11-24 15:25:46 -08:00
Alan Tull
07687c031d fpga manager: remove label
Remove implementation of 'label' DT binding.

Signed-off-by: Alan Tull <atull@opensource.altera.com>
Reviewed-by: Moritz Fischer <moritz.fischer@ettus.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
2015-11-24 15:25:46 -08:00
Moritz Fischer
e2d8680741 fpga: socfpga: Fix check of return value of devm_request_irq
The return value should be checked for non-zero, instead
of checking it being IS_ERR_VALUE().

Acked-by: Alan Tull <atull@opensource.altera.com>
Reviewed-by: Josh Cartwright <joshc@eso.teric.us>
Signed-off-by: Moritz Fischer <moritz.fischer@ettus.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
2015-10-29 15:20:25 -07:00
Moritz Fischer
28f98a12f7 fpga: zynq-fpga: Fix issue with drvdata being overwritten.
Upon registering a FPGA Manager low level driver, FPGA Manager
core overwrites the platform drvdata pointer. Prior to this commit
zynq-fpga falsely relied on this pointer to still be valid at remove()
time.

Reported-by: Alan Tull <atull@opensource.altera.com>
Signed-off-by: Moritz Fischer <moritz.fischer@ettus.com>
Acked-by: Alan Tull <atull@opensource.altera.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
2015-10-23 16:49:44 -07:00
Alan Tull
92d94a7ec5 fpga manager: remove unnecessary null pointer checks
Remove unnecessary null pointer checks.  We want the caller of
these functions to do their own pointer checks.  Add some
comments to document this.

Signed-off-by: Alan Tull <atull@opensource.altera.com>
Reviewed-by: Moritz Fischer <moritz.fischer@ettus.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
2015-10-23 16:49:44 -07:00
Alan Tull
654ba4cc0f fpga manager: ensure lifetime with of_fpga_mgr_get
Ensure device and driver lifetime from of_fpga_mgr_get() to
fpga_mgr_put().

* Don't put_device() in of_fpga_mgr_get, do it in fpga_mgr_put().
  (still do put_device if there is an error).
* Do module_get on the low level driver.
* Don't need to module_get(THIS_MODULE) since we won't be allowed
  to unload the fpga manager core without unloading low level
  driver first.
* Remove unnedessary null check for node pointer.

Signed-off-by: Alan Tull <atull@opensource.altera.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
2015-10-23 16:49:44 -07:00
Moritz Fischer
4d10eaff5b fpga: zynq-fpga: Change fw format to handle bin instead of bit.
This gets rid of the code to strip away the header and byteswap,
as well as the check for the sync word.

Signed-off-by: Moritz Fischer <moritz.fischer@ettus.com>
Reviewed-by: Josh Cartwright <joshc@ni.com>
Acked-by: Michal Simek <michal.simek@xilinx.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
2015-10-23 16:49:12 -07:00
Moritz Fischer
6376931bab fpga: zynq-fpga: Fix unbalanced clock handling
This commit fixes the unbalanced clock handling, where
a failed probe would leave the clock with an enable count of -1.

Reported-by: Josh Cartwright <joshc@ni.com>
Signed-off-by: Moritz Fischer <moritz.fischer@ettus.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
2015-10-23 16:49:12 -07:00
Moritz Fischer
37784706bf fpga manager: Adding FPGA Manager support for Xilinx Zynq 7000
This commit adds FPGA Manager support for the Xilinx Zynq chip.
The code borrows some from the xdevcfg driver in Xilinx'
vendor tree.

Signed-off-by: Moritz Fischer <moritz.fischer@ettus.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
2015-10-17 21:57:16 -07:00
Alan Tull
fab6266e82 fpga manager: add driver for socfpga fpga manager
Add driver to fpga manager framework to allow configuration
of FPGA in Altera SoCFPGA parts.

Signed-off-by: Alan Tull <atull@opensource.altera.com>
Acked-by: Michal Simek <michal.simek@xilinx.com>
Acked-by: Moritz Fischer <moritz.fischer@ettus.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
2015-10-07 18:08:15 +01:00
Alan Tull
6a8c3be7ec add FPGA manager core
API to support programming FPGA's.

The following functions are exported as GPL:
* fpga_mgr_buf_load
   Load fpga from image in buffer

* fpga_mgr_firmware_load
   Request firmware and load it to the FPGA.

* fpga_mgr_register
* fpga_mgr_unregister
   FPGA device drivers can be added by calling
   fpga_mgr_register() to register a set of
   fpga_manager_ops to do device specific stuff.

* of_fpga_mgr_get
* fpga_mgr_put
   Get/put a reference to a fpga manager.

The following sysfs files are created:
* /sys/class/fpga_manager/<fpga>/name
  Name of low level driver.

* /sys/class/fpga_manager/<fpga>/state
  State of fpga manager

Signed-off-by: Alan Tull <atull@opensource.altera.com>
Acked-by: Michal Simek <michal.simek@xilinx.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
2015-10-07 18:08:15 +01:00