linux-stable.git/drivers/spi, branch v4.4.227

spi: dw: use "smp_mb()" to avoid sending spi data error

2020-06-11T07:21:37+00:00

[ Upstream commit bfda044533b213985bc62bd7ca96f2b984d21b80 ]

Because of out-of-order execution about some CPU architecture,
In this debug stage we find Completing spi interrupt enable ->
prodrucing TXEI interrupt -> running "interrupt_transfer" function
will prior to set "dw->rx and dws->rx_end" data, so this patch add
memory barrier to enable dw->rx and dw->rx_end to be visible and
solve to send SPI data error.
eg:
it will fix to this following low possibility error in testing environment
which using SPI control to connect TPM Modules

kernel: tpm tpm0: Operation Timed out
kernel: tpm tpm0: tpm_relinquish_locality: : error -1

Signed-off-by: fengsheng 
Signed-off-by: Xinwei Kong 
Link: https://lore.kernel.org/r/1578019930-55858-1-git-send-email-kong.kongxinwei@hisilicon.com
Signed-off-by: Mark Brown 
Signed-off-by: Sasha Levin

spi: spi-dw: Add lock protect dw_spi rx/tx to prevent concurrent calls

2020-05-20T06:11:43+00:00

commit 19b61392c5a852b4e8a0bf35aecb969983c5932d upstream.

dw_spi_irq() and dw_spi_transfer_one concurrent calls.

I find a panic in dw_writer(): txw = *(u8 *)(dws->tx), when dw->tx==null,
dw->len==4, and dw->tx_end==1.

When tpm driver's message overtime dw_spi_irq() and dw_spi_transfer_one
may concurrent visit dw_spi, so I think dw_spi structure lack of protection.

Otherwise dw_spi_transfer_one set dw rx/tx buffer and then open irq,
store dw rx/tx instructions and other cores handle irq load dw rx/tx
instructions may out of order.

	[ 1025.321302] Call trace:
	...
	[ 1025.321319]  __crash_kexec+0x98/0x148
	[ 1025.321323]  panic+0x17c/0x314
	[ 1025.321329]  die+0x29c/0x2e8
	[ 1025.321334]  die_kernel_fault+0x68/0x78
	[ 1025.321337]  __do_kernel_fault+0x90/0xb0
	[ 1025.321346]  do_page_fault+0x88/0x500
	[ 1025.321347]  do_translation_fault+0xa8/0xb8
	[ 1025.321349]  do_mem_abort+0x68/0x118
	[ 1025.321351]  el1_da+0x20/0x8c
	[ 1025.321362]  dw_writer+0xc8/0xd0
	[ 1025.321364]  interrupt_transfer+0x60/0x110
	[ 1025.321365]  dw_spi_irq+0x48/0x70
	...

Signed-off-by: wuxu.wu 
Link: https://lore.kernel.org/r/1577849981-31489-1-git-send-email-wuxu.wu@huawei.com
Signed-off-by: Mark Brown 
Signed-off-by: Nobuhiro Iwamatsu (CIP) 
Signed-off-by: Greg Kroah-Hartman

spi/zynqmp: remove entry that causes a cs glitch

2020-04-02T17:02:28+00:00

[ Upstream commit 5dd8304981ecffa77bb72b1c57c4be5dfe6cfae9 ]

In the public interface for chipselect, there is always an entry
commented as "Dummy generic FIFO entry" pushed down to the fifo right
after the activate/deactivate command. The dummy entry is 0x0,
irregardless if the intention was to activate or deactive the cs. This
causes the cs line to glitch rather than beeing activated in the case
when there was an activate command.

This has been observed on oscilloscope, and have caused problems for at
least one specific flash device type connected to the qspi port. After
the change the glitch is gone and cs goes active when intended.

The reason why this worked before (except for the glitch) was because
when sending the actual data, the CS bits are once again set. Since
most flashes uses mode 0, there is always a half clk period anyway for
cs to clk active setup time. If someone would rely on timing from a
chip_select call to a transfer_one, it would fail though.

It is unknown why the dummy entry was there in the first place, git log
seems to be of no help in this case. The reference manual gives no
indication of the necessity of this. In fact the lower 8 bits are a
setup (or hold in case of deactivate) time expressed in cycles. So this
should not be needed to fulfill any setup/hold timings.

Signed-off-by: Thommy Jakobsson 
Reviewed-by: Naga Sureshkumar Relli 
Link: https://lore.kernel.org/r/20200224162643.29102-1-thommyj@gmail.com
Signed-off-by: Mark Brown 
Signed-off-by: Sasha Levin

spi: qup: call spi_qup_pm_resume_runtime before suspending

2020-04-02T17:02:27+00:00

[ Upstream commit 136b5cd2e2f97581ae560cff0db2a3b5369112da ]

spi_qup_suspend() will cause synchronous external abort when
runtime suspend is enabled and applied, as it tries to
access SPI controller register while clock is already disabled
in spi_qup_pm_suspend_runtime().

Signed-off-by: Yuji sasaki 
Signed-off-by: Vinod Koul 
Link: https://lore.kernel.org/r/20200214074340.2286170-1-vkoul@kernel.org
Signed-off-by: Mark Brown 
Signed-off-by: Sasha Levin

spi: spi-fsl-spi: call spi_finalize_current_message() at the end

2020-01-29T09:21:44+00:00

[ Upstream commit 44a042182cb1e9f7916e015c836967bf638b33c4 ]

spi_finalize_current_message() shall be called once all
actions are finished, otherwise the last actions might
step over a newly started transfer.

Fixes: c592becbe704 ("spi: fsl-(e)spi: migrate to generic master queueing")
Signed-off-by: Christophe Leroy 
Signed-off-by: Mark Brown 
Signed-off-by: Sasha Levin

spi: bcm2835aux: fix driver to not allow 65535 (=-1) cs-gpios

2020-01-29T09:21:42+00:00

[ Upstream commit 509c583620e9053e43d611bf1614fc3d3abafa96 ]

The original driver by default defines num_chipselects as -1.
This actually allicates an array of 65535 entries in
of_spi_register_master.

There is a side-effect for buggy device trees that (contrary to
dt-binding documentation) have no cs-gpio defined.

This mode was never supported by the driver due to limitations
of native cs and additional code complexity and is explicitly
not stated to be implemented.

To keep backwards compatibility with such buggy DTs we limit
the number of chip_selects to 1, as for all practical purposes
it is only ever realistic to use a single chip select in
native cs mode without negative side-effects.

Fixes: 1ea29b39f4c812ec ("spi: bcm2835aux: add bcm2835 auxiliary spi device...")
Signed-off-by: Martin Sperl 
Acked-by: Stefan Wahren 
Signed-off-by: Mark Brown 
Signed-off-by: Sasha Levin

spi: tegra114: fix for unpacked mode transfers

2020-01-29T09:21:42+00:00

[ Upstream commit 1a89ac5b91895127f7c586ec5075c3753ca25501 ]

Fixes: computation of actual bytes to fill/receive in/from FIFO in unpacked
mode when transfer length is not a multiple of requested bits per word.

unpacked mode transfers fails when the transfer includes partial bytes in
the last word.

Total words to be written/read to/from FIFO is computed based on transfer
length and bits per word. Unpacked mode includes 0 padding bytes for partial
words to align with bits per word and these extra bytes are also accounted
for calculating bytes left to transfer in the current driver.

This causes extra bytes access of tx/rx buffers along with buffer index
position crossing actual length where remain_len becomes negative and due to
unsigned type, negative value is a 32 bit representation of signed value
and transferred bytes never meets the actual transfer length resulting in
transfer timeout and a hang.

This patch fixes this with proper computation of the actual bytes to fill in
FIFO during transmit and the actual bytes to read from FIFO during receive
ignoring 0 padded bytes.

Signed-off-by: Sowjanya Komatineni 
Signed-off-by: Mark Brown 
Signed-off-by: Sasha Levin

spi: tegra114: clear packed bit for unpacked mode

2020-01-29T09:21:42+00:00

[ Upstream commit 7b3d10cdf54b8bc1dc0da21faed9789ac4da3684 ]

Fixes: Clear packed bit when not using packed mode.

Packed bit is not cleared when not using packed mode. This results
in transfer timeouts for the unpacked mode transfers followed by the
packed mode transfers.

Signed-off-by: Sowjanya Komatineni 
Signed-off-by: Mark Brown 
Signed-off-by: Sasha Levin

spi: atmel: fix handling of cs_change set on non-last xfer

2020-01-23T07:18:37+00:00

commit fed8d8c7a6dc2a76d7764842853d81c770b0788e upstream.

The driver does the wrong thing when cs_change is set on a non-last
xfer in a message.  When cs_change is set, the driver deactivates the
CS and leaves it off until a later xfer again has cs_change set whereas
it should be briefly toggling CS off and on again.

This patch brings the behaviour of the driver back in line with the
documentation and common sense.  The delay of 10 us is the same as is
used by the default spi_transfer_one_message() function in spi.c.
[gregory: rebased on for-5.5 from spi tree]
Fixes: 8090d6d1a415 ("spi: atmel: Refactor spi-atmel to use SPI framework queue")
Signed-off-by: Mans Rullgard 
Acked-by: Nicolas Ferre 
Signed-off-by: Gregory CLEMENT 
Link: https://lore.kernel.org/r/20191018153504.4249-1-gregory.clement@bootlin.com
Signed-off-by: Mark Brown 
Signed-off-by: Greg Kroah-Hartman

spi: tegra20-slink: add missed clk_unprepare

2020-01-04T12:34:16+00:00

[ Upstream commit 04358e40ba96d687c0811c21d9dede73f5244a98 ]

The driver misses calling clk_unprepare in probe failure and remove.
Add the calls to fix it.

Signed-off-by: Chuhong Yuan 
Link: https://lore.kernel.org/r/20191115083122.12278-1-hslester96@gmail.com
Signed-off-by: Mark Brown 
Signed-off-by: Sasha Levin