Linux/PROCESSOR-SDK-DRA7X: Ask to proof interrupt configuration

Hello,

The configuration seems correct. You can also check by doing a cat on /sys/kernel/debug/gpio (cat /sys/kernel/debug/gpio) It should aswell show the gpio pin configured as IRQ.
One more thing to verify would be to check the pinmux for GP6[20] is that configured correctly.

Cheers,
--Prabhakar Lad

Hallo,

the GPIO-212 doesn't apper in /sys/kernel/debug/gpio. So I think I have to pinmux this pin. This document (http://www.ti.com/lit/ug/sprui50/sprui50.pdf

says on page 28 in the last column: XREF_CLK3 - GPIO6_20 - (C23). So, I think I have to mux this pin. I haven't go the Pad Config Tool, yet. Could I mux this pin without that? And how? I am using the device tree dra7.dtsi.

Greetings

edit:

Is it right, that I could config the pinmuxes in u-boot? So u-boot is responsible for pinmuxing?

I think in board/ti/dra7xx/mux_data.h there is the right pin:

{XREF_CLK3, (M14 | PIN_INPUT)},		/* xref_clk3.gpio6_20 */

It is defined in arch/arm/include/asm/arch-omap4/mux_dra7xx.h.

The interrupt will be triggered on a rising edge. So, I have to set PIN_INPUT_PULLDOWN?

Which mode I have to set here to use gpio6_20? There are M0 - M15 modes.

Hello,

For GPIO6_20 it needs to be M14, which is being set correct in uboot, try with PIN_INPUT if it doesnt work change it to PIN_INPUT_PULLDOWN.

What is the output of cat /sys/kernel/debug/gpio ? Is the qca,qca7000 module loaded ? Do you get any errors ?

Cheers,
--Prabhakar Lad

Prabhakar Lad said:

For GPIO6_20 it needs to be M14, which is being set correct in uboot, try with PIN_INPUT if it doesnt work change it to PIN_INPUT_PULLDOWN.

Both don't work, Are you sure M14 is right? I thought M14 is for xref_clk3. Is there a documentation for the modes?

Prabhakar Lad said:

What is the output of cat /sys/kernel/debug/gpio ? Is the qca,qca7000 module loaded ? Do you get any errors ?

/sys/kernel/debug/gpio

GPIOs 0-31, platform/4ae10000.gpio, gpio:

GPIOs 32-63, platform/48055000.gpio, gpio:

GPIOs 64-95, platform/48057000.gpio, gpio:

GPIOs 96-127, platform/48059000.gpio, gpio:

GPIOs 128-159, platform/4805b000.gpio, gpio:
 gpio-136 (                    |vmmcwl_fixed        ) out lo    

GPIOs 160-191, platform/4805d000.gpio, gpio:
 gpio-161 (                    |radio_rst           ) out hi    
 gpio-187 (                    |cd                  ) in  lo IRQ

GPIOs 192-223, platform/48051000.gpio, gpio:
 gpio-203 (                    |vtt_fixed           ) out hi    
 gpio-204 (                    |hdmi_hpd            ) in  lo IRQ

GPIOs 224-255, platform/48053000.gpio, gpio:

GPIOs 462-477, i2c/1-0026, pcf8575, can sleep:
 gpio-463 (                    |vin6_sel_s0         ) out lo    
 gpio-466 (                    |hdmi_ct_cp_hpd      ) out hi    
 gpio-467 (                    |hdmi_ls_oe          ) out lo    

GPIOs 478-493, i2c/0-0021, pcf8575, can sleep:
 gpio-479 (                    |id                  ) in  hi IRQ
 gpio-480 (                    |id                  ) in  hi IRQ
 gpio-483 (                    |evm_3v3_sd          ) out hi    

GPIOs 494-509, i2c/0-0020, pcf8575, can sleep:
 gpio-496 (                    |btnUser1            ) in  hi IRQ
 gpio-497 (                    |btnUser2            ) in  hi IRQ
 gpio-498 (                    |?                   ) out hi    
 gpio-499 (                    |?                   ) out hi    
 gpio-500 (                    |?                   ) out hi    
 gpio-501 (                    |?                   ) out hi    

GPIOs 510-511, platform/50000000.gpmc, omap-gpmc:

I think there has to be gpio-212 to appear with in hi IRQ, right?

The modul is loaded:

dmesg | grep qca

[    6.226432] qcaspi spi1.0: ver=0.2.7-i, clkspeed=8000000, burst_len=5000, pluggable=1

lsmod | grep qca

qcaspi                 14970  0

There are errors in dmesg, but I think there are not in dependencie with the IRQ

dmesg | grep err

[    0.691112] omap4_sram_init:Unable to allocate sram needed to handle errata I688
[    0.698806] omap4_sram_init:Unable to get sram pool needed to handle errata I688
[    0.982615] pcf857x: probe of 3-0021 failed with error -121
[    1.346168] pcieport 0000:00:00.0: Signaling PME through PCIe PME interrupt
[    1.377071] fpd3_serdes: probe of 7-002c failed with error -5
[    1.768762] ov1063x: probe of 1-0037 failed with error -121
[    5.902328]  remoteproc0: Direct firmware load for dra7-ipu1-fw.xem4 failed with error -2
[    6.052063]  remoteproc1: Direct firmware load for dra7-ipu2-fw.xem4 failed with error -2
[    6.156329]  remoteproc2: Direct firmware load for dra7-dsp1-fw.xe66 failed with error -2
[    6.295772]  remoteproc3: Direct firmware load for dra7-dsp2-fw.xe66 failed with error -2

I also disabled CONFIG_OMAP_MUX in the Kernel to get the pinmux from u-boot.

Thats what U-Boot says to my board:

CPU  : DRA752-GP ES2.0
Model: TI DRA742
Board: DRA74x EVM REV H.0
DRAM:  4 GiB
MMC:   OMAP SD/MMC: 0, OMAP SD/MMC: 1
SCSI:  SATA link 0 timeout.
AHCI 0001.0300 32 slots 1 ports 3 Gbps 0x1 impl SATA mode
flags: 64bit ncq stag pm led clo only pmp pio slum part ccc apst 

So, I think I am in the right files.

Hello,

1] Referring to [1] page 93, M14 is to select gpio6_20. I didnt try much to find the base address of pinmux, you can use devmem2
and check the value at that address.

2] No it needs to be gpio-180, because base of gpio6 is 160, gpio6_20 will be = 160 + 20 = 180

Example:
GPIOs 160-191, platform/4805d000.gpio, gpio:
gpio-161 ( |radio_rst ) out hi
gpio-180 ( |xx ) in lo IRQ
gpio-187 ( |cd ) in lo IRQ

3] Can you scope the pin gpio6_20 ? If yes comment the following DT node

qca7000@0 {
compatible = "qca,qca7000";
reg = <0>;
interrupt-parent = <&gpio6>;
interrupts = <20 1>;
spi-cpha;
spi-cpol;
spi-max-frequency = <8000000>;
local-mac-address = [ FA DB 12 79 8C 9D ];
/*qca,legacy-mode;*/
};
Then follow [2] to set it as output and change the values from 1/0 and see if desired result is
obtained on the scope

[1] www.ti.com/.../dra746.pdf
[2] elinux.org/GPIO

Cheers,
--Prabhakar Lad

Hallo,

Manual GPIO handling via sysfs with QCA-Node:

root@dra7xx-evm:/sys/class/gpio# echo 180 >> export 
root@dra7xx-evm:/sys/class/gpio# cat gpio180/direction 
in
root@dra7xx-evm:/sys/class/gpio# echo out >> gpio180/direction 
[   45.843756] gpio-180 (sysfs): _gpiod_direction_output_raw: tried to set a GPIO tied to an IRQ as output
-sh: echo: write error: Input/output error

cat /sys/kernel/debug/gpio

GPIOs 160-191, platform/4805d000.gpio, gpio:
 gpio-161 (                    |radio_rst           ) out hi    
 gpio-180 (                    |sysfs               ) in  hi IRQ
 gpio-187 (                    |cd                  ) in  lo IRQ

The IRQ is there but it is hi and not lo and not set by the driver.

Manual GPIO handling via sysfs without QCA-Node:

I can set everyrthing without errors and I can scope the 3,3V on high and 0V on low.

After that I tried to set the GPIO default direction to 'out' from 'in' in the U-Boot source in board/ti/dra7xx/mux_data.h from

{XREF_CLK3, {M14 | PIN_OUTPUT}} to {XREF_CLK3, {M14 | PIN_INPUT}}

to proof, that I am in the right file. In my opinion, if I activate the GPIO 180 (echo 180 >> /sys/class/gpio/export) the direction have to change to 'out' but it is still 'in'.

Is it possible that I'm in the wrong file? The pinmuxing is done by U-Boot?

Greetings

Hallo,

is it possible that the interrupt were not exported to debugfs? Is a subsystem responsible for this? Or is the driver responsible?

In /proc/interrupts the second row (I think it es the counter, now 390) increments by time.

223:        390          0  4805d000.gpio  20 Edge      eth2

So, with this and the post above, is it possible that the interrupt were correct recognized by the kernel, but I don't see this?

Greetings