dt-bindings: cadence-quadspi: add options reset property
Subsystems (3)
PATCH v4 1/2 dt-bindings: cadence-quadspi: add options reset property
The QSPI module can have an optional reset signals that will hold the module in a reset state.
---
v4: no change
v3: created base on review comments
v2: did not exist
v1: did not exist
--- Documentation/devicetree/bindings/mtd/cadence-quadspi.txt | 4 ++++ 1 file changed, 4 insertions(+)
Documentation/devicetree/bindings/mtd/cadence-quadspi.txt +4 -0
@@ -35,6 +35,8 @@ custom properties: (qspi_n_ss_out). - cdns,tslch-ns : Delay in nanoseconds between setting qspi_n_ss_out low and first bit transfer. +- resets : Must contain an entry for each entry in reset-names. + See ../reset/reset.txt for details. Example:
@@ -50,6 +52,8 @@ Example: cdns,fifo-depth = <128>; cdns,fifo-width = <4>; cdns,trigger-address = <0x00000000>; + resets = <&rst QSPI_RESET>, <&rst QSPI_OCP_RESET>; + reset-names = "qspi", "qspi-ocp"; flash0: n25q00@0 { ...
Signature · 1 line
2.20.0
PATCH v4 2/2 mtd: spi-nor: cadence-quadspi: add reset control
Get the reset control properties for the QSPI controller and bring them out of reset. Most will have just one reset bit, but there is an additional OCP reset bit that is used ECC. The OCP reset bit will also need to get de-asserted as well. [1]
---
v4: fix compile error
v3: return full error by using PTR_ERR(rtsc)
move reset control calls until after the clock enables use udelay(2) to be safe
Add optional OCP(Open Core Protocol) reset signal
v2: use devm_reset_control_get_optional_exclusive
print an error message
return -EPROBE_DEFER
--- drivers/mtd/spi-nor/cadence-quadspi.c | 30 +++++++++++++++++++++++++++ 1 file changed, 30 insertions(+)
drivers/mtd/spi-nor/cadence-quadspi.c +30 -0
@@ -34,6 +34,7 @@#include <linux/of.h> #include <linux/platform_device.h> #include <linux/pm_runtime.h> +#include <linux/reset.h> #include <linux/sched.h> #include <linux/spi/spi.h> #include <linux/timer.h>
@@ -1336,6 +1337,8 @@ static int cqspi_probe(struct platform_device *pdev)struct cqspi_st *cqspi;struct resource *res;struct resource *res_ahb; +struct reset_control *rstc; +struct reset_control *rstc_ocp;const struct cqspi_driver_platdata *ddata;int ret;int irq;
@@ -1402,6 +1405,33 @@ static int cqspi_probe(struct platform_device *pdev) goto probe_clk_failed; } +/* Obtain QSPI reset control */ + rstc =devm_reset_control_get_optional_exclusive (dev,"qspi" ); +if (IS_ERR (rstc)) { +dev_err (dev,"Cannot get QSPI reset.\n" ); +if (PTR_ERR (rstc) == -EPROBE_DEFER ) +return PTR_ERR (rstc); + } + + rstc_ocp =devm_reset_control_get_optional_exclusive (dev,"qspi-ocp" ); +if (IS_ERR (rstc_ocp)) { +dev_err (dev,"Cannot get QSPI OCP reset.\n" ); +if (PTR_ERR (rstc_ocp) == -EPROBE_DEFER ) +return PTR_ERR (rstc_ocp); + } + +if (rstc) { +reset_control_assert (rstc); +udelay (2 ); +reset_control_deassert (rstc); + } + +if (rstc_ocp) { +reset_control_assert (rstc_ocp); +udelay (2 ); +reset_control_deassert (rstc_ocp); + } + cqspi->master_ref_clk_hz =clk_get_rate (cqspi->clk); ddata =of_device_get_match_data (dev);if (ddata && (ddata->quirks &CQSPI_NEEDS_WR_DELAY ))
Signature · 1 line
2.20.0
Rob Herring
maintainerQuoted 18 lines · Dinh Nguyen
The QSPI module can have an optional reset signals that will hold the module in a reset state.
Signed-off-by Dinh Nguyen <[email protected]>---
v4: no change
v3: created base on review comments
v2: did not exist
v1: did not exist--- Documentation/devicetree/bindings/mtd/cadence-quadspi.txt | 4 ++++ 1 file changed, 4 insertions(+)Documentation/devicetree/bindings/mtd/cadence-quadspi.txt +0
-0
Documentation/devicetree/bindings/mtd/cadence-quadspi.txt custom properties: · line 39
Quoted 5 lines
Documentation/devicetree/bindings/mtd/cadence-quadspi.txt +2
-0@@ -35,6 +35,8 @@ custom properties: (qspi_n_ss_out). - cdns,tslch-ns : Delay in nanoseconds between setting qspi_n_ss_out low and first bit transfer. +- resets : Must contain an entry for each entry in reset-names. + See ../reset/reset.txt for details.
reset-names needs to be documented with the values and order.
Quoted 15 lines
Documentation/devicetree/bindings/mtd/cadence-quadspi.txt +2
-0Example:@@ -50,6 +52,8 @@ Example: cdns,fifo-depth = <128>; cdns,fifo-width = <4>; cdns,trigger-address = <0x00000000>; + resets = <&rst QSPI_RESET>, <&rst QSPI_OCP_RESET>; + reset-names = "qspi", "qspi-ocp"; flash0: n25q00@0 {Signature · 1 line
2.20.0
Re: [PATCHv4 2/2] mtd: spi-nor: cadence-quadspi: add reset control
In reply to Dinh NguyenQuoted 5 lines · Dinh Nguyen
Get the reset control properties for the QSPI controller and bring them out of reset. Most will have just one reset bit, but there is an additional OCP reset bit that is used ECC. The OCP reset bit will also need to get de-asserted as well. [1]
It's always good to say why the change is needed, e.g. reset the controller at init to have it in a clean state in case the bootloader messed with it.
Quoted 39 lines
Suggested-by Tien-Fong Chee <[email protected]>Signed-off-by Dinh Nguyen <[email protected]>---
v4: fix compile error
v3: return full error by usingPTR_ERR(rtsc)
move reset control calls until after the clock enables use udelay(2) to be safe
Add optional OCP(Open Core Protocol) reset signal
v2: usedevm_reset_control_get_optional_exclusive
print an error message
return -EPROBE_DEFER--- drivers/mtd/spi-nor/cadence-quadspi.c | 30 +++++++++++++++++++++++++++ 1 file changed, 30 insertions(+)drivers/mtd/spi-nor/cadence-quadspi.c +3
-0@@ -34,6 +34,7 @@#include <linux/of.h> #include <linux/platform_device.h> #include <linux/pm_runtime.h> +#include <linux/reset.h> #include <linux/sched.h> #include <linux/spi/spi.h> #include <linux/timer.h> @@ -1336,6 +1337,8 @@ static int cqspi_probe(struct platform_device *pdev)struct cqspi_st *cqspi;struct resource *res;struct resource *res_ahb; +struct reset_control *rstc; +struct reset_control *rstc_ocp;const struct cqspi_driver_platdata *ddata;int ret;int irq;
drivers/mtd/spi-nor/cadence-quadspi.c static int cqspi_probe(struct platform_device *pdev) · line 1412
Quoted 8 lines
drivers/mtd/spi-nor/cadence-quadspi.c +5
-0@@ -1402,6 +1405,33 @@ static int cqspi_probe(struct platform_device *pdev) goto probe_clk_failed; } +/* Obtain QSPI reset control */ + rstc =devm_reset_control_get_optional_exclusive (dev,"qspi" ); +if (IS_ERR (rstc)) { +dev_err (dev,"Cannot get QSPI reset.\n" ); +if (PTR_ERR (rstc) == -EPROBE_DEFER )
what I meant was to get rid of this if and return PTR_ERR(rstc) directly.
drivers/mtd/spi-nor/cadence-quadspi.c static int cqspi_probe(struct platform_device *pdev) · line 1425
Quoted 12 lines
drivers/mtd/spi-nor/cadence-quadspi.c +12
-0+return PTR_ERR (rstc); + } + + rstc_ocp =devm_reset_control_get_optional_exclusive (dev,"qspi-ocp" ); +if (IS_ERR (rstc_ocp)) { +dev_err (dev,"Cannot get QSPI OCP reset.\n" ); +if (PTR_ERR (rstc_ocp) == -EPROBE_DEFER ) +return PTR_ERR (rstc_ocp); + } + + if (rstc) {> + reset_control_assert(rstc); +udelay (2 );
why 2us? what's the appropriate length of time that we should wait between assert and deassert?
drivers/mtd/spi-nor/cadence-quadspi.c static int cqspi_probe(struct platform_device *pdev) · line 1430
Quoted 5 lines
drivers/mtd/spi-nor/cadence-quadspi.c +5
-0+reset_control_deassert (rstc); + } + +if (rstc_ocp) { +reset_control_assert (rstc_ocp);
Does it mater the order in which you assert these signals? can we group these module resets asserts, i.e. first do the assert for both rstc and rstcp and then the deassert?
drivers/mtd/spi-nor/cadence-quadspi.c static int cqspi_probe(struct platform_device *pdev) · line 1431
Quoted 2 lines
drivers/mtd/spi-nor/cadence-quadspi.c +2
-0+udelay (2 ); +reset_control_deassert (rstc_ocp);
Is software deassert needed? I'm looking at [2], Table 46. PER1 Group, Generated Module Resets, and it seems that software deassert is not an option for qspi_flash_ecc_rst_n
[2]https://www.intel.com/content/dam/www/programmable/us/en/pdfs/literature/hb/arria-10/a10_5v4.pdf
Quoted 6 lines
drivers/mtd/spi-nor/cadence-quadspi.c +2
-0+ } + cqspi->master_ref_clk_hz =clk_get_rate (cqspi->clk); ddata =of_device_get_match_data (dev);if (ddata && (ddata->quirks &CQSPI_NEEDS_WR_DELAY ))
Re: [PATCHv4 2/2] mtd: spi-nor: cadence-quadspi: add reset control
Quoted 11 lines · On 6/6/19 3:26 AM, [email protected] wrote:
Quoted 5 lines · Dinh Nguyen
Get the reset control properties for the QSPI controller and bring them out of reset. Most will have just one reset bit, but there is an additional OCP reset bit that is used ECC. The OCP reset bit will also need to get de-asserted as well. [1]
It's always good to say why the change is needed, e.g. reset the controller at init to have it in a clean state in case the bootloader messed with it.
Will update..
Quoted 43 lines
Quoted 43 lines
Suggested-by Tien-Fong Chee <[email protected]>Signed-off-by Dinh Nguyen <[email protected]>---
v4: fix compile error
v3: return full error by usingPTR_ERR(rtsc)
move reset control calls until after the clock enables use udelay(2) to be safe
Add optional OCP(Open Core Protocol) reset signal
v2: usedevm_reset_control_get_optional_exclusive
print an error message
return -EPROBE_DEFER--- drivers/mtd/spi-nor/cadence-quadspi.c | 30 +++++++++++++++++++++++++++ 1 file changed, 30 insertions(+)drivers/mtd/spi-nor/cadence-quadspi.c +3
-0@@ -34,6 +34,7 @@#include <linux/of.h> #include <linux/platform_device.h> #include <linux/pm_runtime.h> +#include <linux/reset.h> #include <linux/sched.h> #include <linux/spi/spi.h> #include <linux/timer.h> @@ -1336,6 +1337,8 @@ static int cqspi_probe(struct platform_device *pdev)struct cqspi_st *cqspi;struct resource *res;struct resource *res_ahb; +struct reset_control *rstc; +struct reset_control *rstc_ocp;const struct cqspi_driver_platdata *ddata;int ret;int irq;@@ -1402,6 +1405,33 @@ static int cqspi_probe(struct platform_device *pdev) goto probe_clk_failed; }
Quoted 8 lines
drivers/mtd/spi-nor/cadence-quadspi.c static int cqspi_probe(struct platform_device *pdev) · line 1412
Quoted 5 lines
drivers/mtd/spi-nor/cadence-quadspi.c +5
-0+/* Obtain QSPI reset control */ + rstc =devm_reset_control_get_optional_exclusive (dev,"qspi" ); +if (IS_ERR (rstc)) { +dev_err (dev,"Cannot get QSPI reset.\n" ); +if (PTR_ERR (rstc) == -EPROBE_DEFER )what I meant was to get rid of this if and return
PTR_ERR(rstc)directly.
Okay..
Quoted 7 lines
Quoted 7 lines
drivers/mtd/spi-nor/cadence-quadspi.c +7
-0+return PTR_ERR (rstc); + } + + rstc_ocp =devm_reset_control_get_optional_exclusive (dev,"qspi-ocp" ); +if (IS_ERR (rstc_ocp)) { +dev_err (dev,"Cannot get QSPI OCP reset.\n" ); +if (PTR_ERR (rstc_ocp) == -EPROBE_DEFER )
Quoted 9 lines
drivers/mtd/spi-nor/cadence-quadspi.c static int cqspi_probe(struct platform_device *pdev) · line 1425
Quoted 5 lines
drivers/mtd/spi-nor/cadence-quadspi.c +5
-0+return PTR_ERR (rstc_ocp); + } + + if (rstc) {> + reset_control_assert(rstc); +udelay (2 );why 2us? what's the appropriate length of time that we should wait between assert and deassert?
This length hasn't been documented anywhere. I've tested with both 2us and none, and both cases seem to be working fine. 2us was something I saw in the STM32 driver. I'll remove it.
Quoted 10 lines
drivers/mtd/spi-nor/cadence-quadspi.c static int cqspi_probe(struct platform_device *pdev) · line 1430
Quoted 5 lines
drivers/mtd/spi-nor/cadence-quadspi.c +5
-0+reset_control_deassert (rstc); + } + +if (rstc_ocp) { +reset_control_assert (rstc_ocp);Does it mater the order in which you assert these signals? can we group these module resets asserts, i.e. first do the assert for both rstc and rstcp and then the deassert?
Quoted 8 lines
drivers/mtd/spi-nor/cadence-quadspi.c static int cqspi_probe(struct platform_device *pdev) · line 1431
Quoted 2 lines
drivers/mtd/spi-nor/cadence-quadspi.c +2
-0+udelay (2 ); +reset_control_deassert (rstc_ocp);Is software deassert needed? I'm looking at [2], Table 46. PER1 Group, Generated Module Resets, and it seems that software deassert is not an option for
qspi_flash_ecc_rst_n[2]https://www.intel.com/content/dam/www/programmable/us/en/pdfs/literature/hb/arria-10/a10_5v4.pdf
I believe this is a mistake. QSPI is not working for me if I don't do a software reset deassert on the ocp bit.
Dinh