linux-toradex.git/drivers/dma/xilinx, branch v6.19

dmaengine: xilinx_dma: Fix uninitialized addr_width when "xlnx,addrwidth" property is missing

2025-12-16T15:46:59+00:00

When device tree lacks optional "xlnx,addrwidth" property, the addr_width
variable remained uninitialized with garbage values, causing incorrect
DMA mask configuration and subsequent probe failure. The fix ensures a
fallback to the default 32-bit address width when this property is missing.

Signed-off-by: Suraj Gupta 
Fixes: b72db4005fe4 ("dmaengine: vdma: Add 64 bit addressing support to the driver")
Reviewed-by: Radhey Shyam Pandey 
Reviewed-by: Folker Schwesinger 
Link: https://patch.msgid.link/20251021183006.3434495-1-suraj.gupta2@amd.com
Signed-off-by: Vinod Koul

dmaengine: xilinx: xdma: Fix regmap max_register

2025-12-14T09:17:23+00:00

The max_register field is assigned the size of the register memory
region instead of the offset of the last register.
The result is that reading from the regmap via debugfs can cause
a segmentation fault:

tail /sys/kernel/debug/regmap/xdma.1.auto/registers
Unable to handle kernel paging request at virtual address ffff800082f70000
Mem abort info:
  ESR = 0x0000000096000007
  EC = 0x25: DABT (current EL), IL = 32 bits
  SET = 0, FnV = 0
  EA = 0, S1PTW = 0
  FSC = 0x07: level 3 translation fault
[...]
Call trace:
 regmap_mmio_read32le+0x10/0x30
 _regmap_bus_reg_read+0x74/0xc0
 _regmap_read+0x68/0x198
 regmap_read+0x54/0x88
 regmap_read_debugfs+0x140/0x380
 regmap_map_read_file+0x30/0x48
 full_proxy_read+0x68/0xc8
 vfs_read+0xcc/0x310
 ksys_read+0x7c/0x120
 __arm64_sys_read+0x24/0x40
 invoke_syscall.constprop.0+0x64/0x108
 do_el0_svc+0xb0/0xd8
 el0_svc+0x38/0x130
 el0t_64_sync_handler+0x120/0x138
 el0t_64_sync+0x194/0x198
Code: aa1e03e9 d503201f f9400000 8b214000 (b9400000)
---[ end trace 0000000000000000 ]---
note: tail[1217] exited with irqs disabled
note: tail[1217] exited with preempt_count 1
Segmentation fault

Fixes: 17ce252266c7 ("dmaengine: xilinx: xdma: Add xilinx xdma driver")
Reviewed-by: Lizhi Hou 
Reviewed-by: Radhey Shyam Pandey 
Reviewed-by: Alexander Stein 
Signed-off-by: Anthony Brandon 
Signed-off-by: Vinod Koul

dmaengine: xilinx_dma: Support descriptor setup from dma_vecs

2025-09-02T07:04:19+00:00

The DMAEngine provides an interface for obtaining DMA transaction
descriptors from an array of scatter gather buffers represented by
struct dma_vec. This interface is used in the DMABUF API of the IIO
framework [1][2].
To enable DMABUF support through the IIO framework for the Xilinx DMA,
implement callback .device_prep_peripheral_dma_vec() of struct
dma_device in the driver.

[1]: 7a86d469983a ("iio: buffer-dmaengine: Support new DMABUF based userspace API")
[2]: 5878853fc938 ("dmaengine: Add API function dmaengine_prep_peripheral_dma_vec()")

Signed-off-by: Folker Schwesinger 
Reviewed-by: Suraj Gupta 
Reviewed-by: Radhey Shyam Pandey 
Link: https://lore.kernel.org/r/DCCKQLKOZC06.2H6LJ8RJQJNV2@folker-schwesinger.de
Signed-off-by: Vinod Koul

dmaengine: zynqmp_dma: Add shutdown operation support

2025-08-20T17:27:20+00:00

Add shutdown callback to ensure that DMA operations are properly stopped
and resources are released during system shutdown or kexec operations.
Fix incorrect PM state handling in the remove function that was causing
clock disable warnings during the shutdown operations, which was not
implemented earlier. The original logic used pm_runtime_enabled() check
after calling the pm_runtime_disable(), would always evaluate to true
after the disable call, which leads to unconditionally calling the
runtime_suspend regardless of the device's actual power state.

During shutdown, the device may already be suspended with clock disabled
from the autosuspend timer, causing the clock framework to warn about
the double-disable attempt. The pm_runtime_active() function checks the
actual device power state rather than the PM subsystem's enabled/disabled
status. ensuring the runtime_suspend is only called when the device is in
active power state. This prevents clock warnings during shutdown while
maintaining proper cleanup during normal remove operations.

Signed-off-by: Abin Joseph 
Reviewed-by: Radhey Shyam Pandey 
Link: https://lore.kernel.org/r/20250722070255.28944-1-abin.joseph@amd.com
Signed-off-by: Vinod Koul

dmaengine: xilinx_dma: Set dma_device directions

2025-05-14T13:50:07+00:00

Coalesce the direction bits from the enabled TX and/or RX channels into
the directions bit mask of dma_device. Without this mask set,
dma_get_slave_caps() in the DMAEngine fails, which prevents the driver
from being used with an IIO DMAEngine buffer.

Signed-off-by: Thomas Gessler 
Reviewed-by: Suraj Gupta 
Tested-by: Folker Schwesinger 
Link: https://lore.kernel.org/r/20250507182101.909010-1-thomas.gessler@brueckmann-gmbh.de
Signed-off-by: Vinod Koul

dmaengine: xilinx_dma: Set max segment size

2025-04-17T15:15:20+00:00

Set the maximumg DMA segment size from the actual core configuration
value. Without this setting, the default value of 64 KiB is reported,
and larger sizes cannot be used for IIO DMAEngine buffers.

Signed-off-by: Thomas Gessler 
Tested-by: Folker Schwesinger 
Link: https://lore.kernel.org/r/20250314134717.703287-1-thomas.gessler@brueckmann-gmbh.de
Signed-off-by: Vinod Koul

dmaengine: Use str_enable_disable-like helpers

2025-02-10T14:27:11+00:00

Replace ternary (condition ? "enable" : "disable") syntax with helpers
from string_choices.h because:
1. Simple function call with one argument is easier to read.  Ternary
   operator has three arguments and with wrapping might lead to quite
   long code.
2. Is slightly shorter thus also easier to read.
3. It brings uniformity in the text - same string.
4. Allows deduping by the linker, which results in a smaller binary
   file.

Signed-off-by: Krzysztof Kozlowski 
Reviewed-by: Manivannan Sadhasivam  #dw-edma
Reviewed-by: Jernej Skrabec 
Link: https://lore.kernel.org/r/20250114191021.854080-1-krzysztof.kozlowski@linaro.org
Signed-off-by: Vinod Koul

dmaengine: xilinx_dma: Configure parking registers only if parking enabled

2024-12-02T17:00:41+00:00

The VDMA can work in two modes, parking or circular. Do not program the
parking mode registers in case circular mode is used, it is useless.

Signed-off-by: Marek Vasut 
Reviewed-by: Radhey Shyam Pandey 
Link: https://lore.kernel.org/r/20241031171132.56861-1-marex@denx.de
Signed-off-by: Vinod Koul

dmaengine: xilinx: xdma: remove redundant check on ret

2024-12-02T16:56:37+00:00

The variable ret is being checked for an error and returning ret
and the following statement returns ret too. The if check is
redundant, and remove it. Just return the value returned from
the call to regmap_write.

Signed-off-by: Colin Ian King 
Link: https://lore.kernel.org/r/20241107114656.17611-1-colin.i.king@gmail.com
Signed-off-by: Vinod Koul

dmaengine: fix typo in the comment

2024-10-21T17:37:08+00:00

Correctly spelled comments make it easier for the reader to understand
the code.

Replace 'enngine' with 'engine' in the comment &
replace 'trascatioin' with 'transaction' in the comment &
replace 'descripter' with 'descriptor' in the comment &
replace 'descritpor' with 'descriptor' in the comment &
replace 'rgisters' with 'registers' in the comment.

Signed-off-by: Yan Zhen 
Acked-by: Palmer Dabbelt 
Link: https://lore.kernel.org/r/20240918034114.860132-1-yanzhen@vivo.com
Signed-off-by: Vinod Koul