linux-stable.git/drivers/soc/qcom, branch v5.15

soc: qcom: mdt_loader: Drop PT_LOAD check on hash segment

2021-09-20T00:55:20+00:00

PT_LOAD type denotes that the segment should be loaded into the final
firmware memory region.  Hash segment is not one such, because it's only
needed for PAS init and shouldn't be in the final firmware memory region.
That's why mdt_phdr_valid() explicitly reject non PT_LOAD segment and
hash segment.  This actually makes the hash segment type check in
qcom_mdt_read_metadata() unnecessary and redundant.  For a hash segment,
it won't be loaded into firmware memory region anyway, due to the
QCOM_MDT_TYPE_HASH check in mdt_phdr_valid(), even if it has a PT_LOAD
type for some reason (misusing or abusing?).

Some firmware files on Sony phones are such examples, e.g WCNSS firmware
of Sony Xperia M4 Aqua phone.  The type of hash segment is just PT_LOAD.
Drop the unnecessary hash segment type check in qcom_mdt_read_metadata()
to fix firmware loading failure on these phones, while hash segment is
still kept away from the final firmware memory region.

Fixes: 498b98e93900 ("soc: qcom: mdt_loader: Support loading non-split images")
Signed-off-by: Shawn Guo 
Reviewed-by: Marijn Suijten 
Signed-off-by: Bjorn Andersson 
Link: https://lore.kernel.org/r/20210828070202.7033-1-shawn.guo@linaro.org

soc: qcom: socinfo: Fixed argument passed to platform_set_data()

2021-09-13T15:39:59+00:00

Set qcom_socinfo pointer as data being stored instead of pointer
to soc_device structure. Aligns with future calls to platform_get_data()
which expects qcom_socinfo pointer.

Fixes: efb448d0a3fc ("soc: qcom: Add socinfo driver")
Signed-off-by: Antonio Martorana 
Reviewed-by: Bjorn Andersson 
Link: https://lore.kernel.org/r/1629159879-95777-1-git-send-email-amartora@codeaurora.org
Signed-off-by: Bjorn Andersson

Merge tag 'drivers-5.15' of git://git.kernel.org/pub/scm/linux/kernel/git/soc/soc

2021-09-01T22:25:28+00:00

Pull ARM SoC driver updates from Arnd Bergmann:
 "These are updates for drivers that are tied to a particular SoC,
  including the correspondig device tree bindings:

   - A couple of reset controller changes for unisoc, uniphier, renesas
     and zte platforms

   - memory controller driver fixes for omap and tegra

   - Rockchip io domain driver updates

   - Lots of updates for qualcomm platforms, mostly touching their
     firmware and power management drivers

   - Tegra FUSE and firmware driver updateѕ

   - Support for virtio transports in the SCMI firmware framework

   - cleanup of ixp4xx drivers, towards enabling multiplatform support
     and bringing it up to date with modern platforms

   - Minor updates for keystone, mediatek, omap, renesas"

* tag 'drivers-5.15' of git://git.kernel.org/pub/scm/linux/kernel/git/soc/soc: (96 commits)
  reset: simple: remove ZTE details in Kconfig help
  soc: rockchip: io-domain: Remove unneeded semicolon
  soc: rockchip: io-domain: add rk3568 support
  dt-bindings: power: add rk3568-pmu-io-domain support
  bus: ixp4xx: return on error in ixp4xx_exp_probe()
  soc: renesas: Prefer memcpy() over strcpy()
  firmware: tegra: Stop using seq_get_buf()
  soc/tegra: fuse: Enable fuse clock on suspend for Tegra124
  soc/tegra: fuse: Add runtime PM support
  soc/tegra: fuse: Clear fuse->clk on driver probe failure
  soc/tegra: pmc: Prevent racing with cpuilde driver
  soc/tegra: bpmp: Remove unused including 
  dt-bindings: soc: ti: pruss: Add dma-coherent property
  soc: ti: Remove pm_runtime_irq_safe() usage for smartreflex
  soc: ti: pruss: Enable support for ICSSG subsystems on K3 AM64x SoCs
  dt-bindings: soc: ti: pruss: Update bindings for K3 AM64x SoCs
  firmware: arm_scmi: Use WARN_ON() to check configured transports
  firmware: arm_scmi: Fix boolconv.cocci warnings
  soc: mediatek: mmsys: Fix missing UFOE component in mt8173 table routing
  soc: mediatek: mmsys: add MT8365 support
  ...

soc: qcom: smsm: Fix missed interrupts if state changes while masked

2021-08-05T03:20:09+00:00

The SMSM driver detects interrupt edges by tracking the last state
it has seen (and has triggered the interrupt handler for). This works
fine, but only if the interrupt does not change state while masked.

For example, if an interrupt is unmasked while the state is HIGH,
the stored last_value for that interrupt might still be LOW. Then,
when the remote processor triggers smsm_intr() we assume that nothing
has changed, even though the state might have changed from HIGH to LOW.

Attempt to fix this by checking the current remote state before
unmasking an IRQ. Use atomic operations to avoid the interrupt handler
from interfering with the unmask function.

This fixes modem crashes in some edge cases with the BAM-DMUX driver.
Specifically, the BAM-DMUX interrupt handler is not called for the
HIGH -> LOW smsm state transition if the BAM-DMUX driver is loaded
(and therefore unmasks the interrupt) after the modem was already started:

qcom-q6v5-mss 4080000.remoteproc: fatal error received: a2_task.c:3188:
  Assert FALSE failed: A2 DL PER deadlock timer expired waiting for Apps ACK

Fixes: c97c4090ff72 ("soc: qcom: smsm: Add driver for Qualcomm SMSM")
Signed-off-by: Stephan Gerhold 
Link: https://lore.kernel.org/r/20210712135703.324748-2-stephan@gerhold.net
Signed-off-by: Bjorn Andersson

soc: qcom: smsm: Implement support for get_irqchip_state

2021-08-05T03:20:03+00:00

At the moment there is no way for drivers to get the current state
of the interrupt signal reported by a remote processor. The irqchip
API has generic functionality for this, using irq_get_irqchip_state().

Implement support for getting the IRQCHIP_STATE_LINE_LEVEL by reading
the remote state and checking the bit for the specified IRQ.

Signed-off-by: Stephan Gerhold 
Link: https://lore.kernel.org/r/20210712135703.324748-1-stephan@gerhold.net
Signed-off-by: Bjorn Andersson

soc: qcom: mdt_loader: be more informative on errors

2021-08-05T03:01:56+00:00

In __qcom_mdt_load() there are cases where an error occurs that
cause a message to be printed.  In some of those cases the errno
value can be helpful to understand exactly what caused the problem.

Print the errno (as well as the firmware file name) where it is
helpful, and in a few cases reword the error message.  Consistently
use the private fw_name for the file name.

Signed-off-by: Alex Elder 
Link: https://lore.kernel.org/r/20210315144455.1770492-1-elder@linaro.org
Signed-off-by: Bjorn Andersson

soc: qcom: smd-rpm: Add SM6115 compatible

2021-08-04T23:22:22+00:00

Add a compatible for SM6115

Signed-off-by: Iskren Chernev 
Link: https://lore.kernel.org/r/20210731164827.2756798-3-iskren.chernev@gmail.com
Signed-off-by: Bjorn Andersson

soc: qcom: geni: Add support for gpi dma

2021-08-04T19:23:31+00:00

GPI DMA is one of the DMA modes supported on geni, this adds support to
enable that mode

Also do better documentation of the enum geni_se_xfer_mode.

Signed-off-by: Vinod Koul 
Reviewed-by: Douglas Anderson 
Link: https://lore.kernel.org/r/20210625052213.32260-3-vkoul@kernel.org
Signed-off-by: Bjorn Andersson

soc: qcom: geni: move GENI_IF_DISABLE_RO to common header

2021-08-04T19:23:27+00:00

GENI_IF_DISABLE_RO is used by geni spi driver as well to check the
status if GENI, so move this to common header qcom-geni-se.h

Reviewed-by: Bjorn Andersson 
Signed-off-by: Vinod Koul 
Reviewed-by: Douglas Anderson 
Link: https://lore.kernel.org/r/20210625052213.32260-2-vkoul@kernel.org
Signed-off-by: Bjorn Andersson

PM: AVS: qcom-cpr: Use nvmem_cell_read_variable_le_u32()

2021-08-03T23:00:18+00:00

Let's delete the private function cpr_read_efuse() since it does the
basically the same thing as the new API call
nvmem_cell_read_variable_le_u32().

Differences between the new API call and the old private function:
* less error printing (I assume this is OK).
* will give an error if the value doesn't fit in 32-bits (the old code
  would have truncated silently).

Reviewed-by: Bjorn Andersson 
Acked-by: Niklas Cassel 
Signed-off-by: Douglas Anderson 
Link: https://lore.kernel.org/r/20210521134437.v2.1.Id1c70158722750aec0673d60c12e46a9c66bbfed@changeid
Signed-off-by: Bjorn Andersson