[RESEND][Patch v4] net: pfe_eth: Use spi_flash_read API to access flash memory
Current PFE firmware access spi-nor memory directly. New spi-mem framework does not support direct memory access. So, let's use spi_flash_read API to access memory instead of directly using it.
Signed-off-by: Kuldeep Singh kuldeep.singh@nxp.com Reviewed-by: Frieder Schrempf frieder.schrempf@kontron.de --- v4: -Rebase to top -Use complete firmware size -Return -ENODEV if flash probe fails v3: -Replace ret with 0 if flash probe fails v2: -Add return error code -Changes in displayed error log
drivers/net/pfe_eth/pfe_firmware.c | 45 +++++++++++++++++++++++++++++++++++++- include/configs/ls1012a_common.h | 5 ++++- 2 files changed, 48 insertions(+), 2 deletions(-)
diff --git a/drivers/net/pfe_eth/pfe_firmware.c b/drivers/net/pfe_eth/pfe_firmware.c index 0493cfe..55e661c 100644 --- a/drivers/net/pfe_eth/pfe_firmware.c +++ b/drivers/net/pfe_eth/pfe_firmware.c @@ -16,13 +16,14 @@ #include <linux/bitops.h> #include <net/pfe_eth/pfe_eth.h> #include <net/pfe_eth/pfe_firmware.h> +#include <spi_flash.h> #ifdef CONFIG_CHAIN_OF_TRUST #include <fsl_validate.h> #endif
#define PFE_FIRMWARE_FIT_CNF_NAME "config@1"
-static const void *pfe_fit_addr = (void *)CONFIG_SYS_LS_PFE_FW_ADDR; +static const void *pfe_fit_addr;
/* * PFE elf firmware loader. @@ -163,6 +164,44 @@ static int pfe_fit_check(void) return ret; }
+int pfe_spi_flash_init(void) +{ + struct spi_flash *pfe_flash; + int ret = 0; + void *addr = malloc(CONFIG_SYS_QE_FMAN_FW_LENGTH); + +#ifdef CONFIG_DM_SPI_FLASH + struct udevice *new; + + /* speed and mode will be read from DT */ + ret = spi_flash_probe_bus_cs(CONFIG_ENV_SPI_BUS, + CONFIG_ENV_SPI_CS, 0, 0, &new); + + pfe_flash = dev_get_uclass_priv(new); +#else + pfe_flash = spi_flash_probe(CONFIG_ENV_SPI_BUS, + CONFIG_ENV_SPI_CS, + CONFIG_ENV_SPI_MAX_HZ, + CONFIG_ENV_SPI_MODE); +#endif + if (!pfe_flash) { + printf("SF: probe for pfe failed\n"); + return -ENODEV; + } + + ret = spi_flash_read(pfe_flash, + CONFIG_SYS_LS_PFE_FW_ADDR, + CONFIG_SYS_QE_FMAN_FW_LENGTH, + addr); + if (ret) + printf("SF: read for pfe failed\n"); + + pfe_fit_addr = addr; + spi_flash_free(pfe_flash); + + return ret; +} + /* * PFE firmware initialization. * Loads different firmware files from FIT image. @@ -187,6 +226,10 @@ int pfe_firmware_init(void) int ret = 0; int fw_count;
+ ret = pfe_spi_flash_init(); + if (ret) + goto err; + ret = pfe_fit_check(); if (ret) goto err; diff --git a/include/configs/ls1012a_common.h b/include/configs/ls1012a_common.h index 3bea9a9..06af8bf 100644 --- a/include/configs/ls1012a_common.h +++ b/include/configs/ls1012a_common.h @@ -36,8 +36,11 @@ /* Size of malloc() pool */ #define CONFIG_SYS_MALLOC_LEN (CONFIG_ENV_SIZE + 1024 * 1024)
-/*SPI device */ +/* PFE */ #define CONFIG_SYS_FMAN_FW_ADDR 0x400d0000 +#define CONFIG_SYS_QE_FMAN_FW_LENGTH 0x300000 + +/*SPI device */ #define CONFIG_SYS_FSL_QSPI_BASE 0x40000000
/* SATA */
On 28.05.20 08:12, Kuldeep Singh wrote:
Current PFE firmware access spi-nor memory directly. New spi-mem framework does not support direct memory access. So, let's use spi_flash_read API to access memory instead of directly using it.
Signed-off-by: Kuldeep Singh kuldeep.singh@nxp.com Reviewed-by: Frieder Schrempf frieder.schrempf@kontron.de
So this patch has been floating around for about half a year now with almost no attention from the maintainers! Several pings have been sent by the author without response.
In fact this patch was blocking 91afd36f3802 ("spi: Transform the FSL QuadSPI driver to use the SPI MEM API") that has finally been merged, ignoring the fact that it will break ls1012a without this being applied too.
I don't want to blame anyone, but I want to voice my disappointment and raise attention that something is obviously wrong here.
v4: -Rebase to top -Use complete firmware size -Return -ENODEV if flash probe fails v3: -Replace ret with 0 if flash probe fails v2: -Add return error code -Changes in displayed error log
drivers/net/pfe_eth/pfe_firmware.c | 45 +++++++++++++++++++++++++++++++++++++- include/configs/ls1012a_common.h | 5 ++++- 2 files changed, 48 insertions(+), 2 deletions(-)
diff --git a/drivers/net/pfe_eth/pfe_firmware.c b/drivers/net/pfe_eth/pfe_firmware.c index 0493cfe..55e661c 100644 --- a/drivers/net/pfe_eth/pfe_firmware.c +++ b/drivers/net/pfe_eth/pfe_firmware.c @@ -16,13 +16,14 @@ #include <linux/bitops.h> #include <net/pfe_eth/pfe_eth.h> #include <net/pfe_eth/pfe_firmware.h> +#include <spi_flash.h> #ifdef CONFIG_CHAIN_OF_TRUST #include <fsl_validate.h> #endif
#define PFE_FIRMWARE_FIT_CNF_NAME "config@1"
-static const void *pfe_fit_addr = (void *)CONFIG_SYS_LS_PFE_FW_ADDR; +static const void *pfe_fit_addr;
/*
- PFE elf firmware loader.
@@ -163,6 +164,44 @@ static int pfe_fit_check(void) return ret; }
+int pfe_spi_flash_init(void) +{
- struct spi_flash *pfe_flash;
- int ret = 0;
- void *addr = malloc(CONFIG_SYS_QE_FMAN_FW_LENGTH);
+#ifdef CONFIG_DM_SPI_FLASH
- struct udevice *new;
- /* speed and mode will be read from DT */
- ret = spi_flash_probe_bus_cs(CONFIG_ENV_SPI_BUS,
CONFIG_ENV_SPI_CS, 0, 0, &new);- pfe_flash = dev_get_uclass_priv(new);
+#else
- pfe_flash = spi_flash_probe(CONFIG_ENV_SPI_BUS,
CONFIG_ENV_SPI_CS,CONFIG_ENV_SPI_MAX_HZ,CONFIG_ENV_SPI_MODE);+#endif
- if (!pfe_flash) {
printf("SF: probe for pfe failed\n");return -ENODEV;- }
- ret = spi_flash_read(pfe_flash,
CONFIG_SYS_LS_PFE_FW_ADDR,CONFIG_SYS_QE_FMAN_FW_LENGTH,addr);- if (ret)
printf("SF: read for pfe failed\n");- pfe_fit_addr = addr;
- spi_flash_free(pfe_flash);
- return ret;
+}
- /*
- PFE firmware initialization.
- Loads different firmware files from FIT image.
@@ -187,6 +226,10 @@ int pfe_firmware_init(void) int ret = 0; int fw_count;
- ret = pfe_spi_flash_init();
- if (ret)
goto err;- ret = pfe_fit_check(); if (ret) goto err;
diff --git a/include/configs/ls1012a_common.h b/include/configs/ls1012a_common.h index 3bea9a9..06af8bf 100644 --- a/include/configs/ls1012a_common.h +++ b/include/configs/ls1012a_common.h @@ -36,8 +36,11 @@ /* Size of malloc() pool */ #define CONFIG_SYS_MALLOC_LEN (CONFIG_ENV_SIZE + 1024 * 1024)
-/*SPI device */ +/* PFE */ #define CONFIG_SYS_FMAN_FW_ADDR 0x400d0000 +#define CONFIG_SYS_QE_FMAN_FW_LENGTH 0x300000
+/*SPI device */ #define CONFIG_SYS_FSL_QSPI_BASE 0x40000000
/* SATA */
-----Original Message----- From: Schrempf Frieder frieder.schrempf@kontron.de Sent: Thursday, May 28, 2020 1:46 PM To: Kuldeep Singh kuldeep.singh@nxp.com; Joe Hershberger joe.hershberger@ni.com; u-boot@lists.denx.de Cc: Priyanka Jain priyanka.jain@nxp.com; Tom Rini trini@konsulko.com Subject: [EXT] Re: [RESEND][Patch v4] net: pfe_eth: Use spi_flash_read API to access flash memory
Caution: EXT Email
On 28.05.20 08:12, Kuldeep Singh wrote:
Current PFE firmware access spi-nor memory directly. New spi-mem framework does not support direct memory access. So, let's use spi_flash_read API to access memory instead of directly using it.
Signed-off-by: Kuldeep Singh kuldeep.singh@nxp.com Reviewed-by: Frieder Schrempf frieder.schrempf@kontron.de
So this patch has been floating around for about half a year now with almost no attention from the maintainers! Several pings have been sent by the author without response.
In fact this patch was blocking 91afd36f3802 ("spi: Transform the FSL QuadSPI driver to use the SPI MEM API") that has finally been merged, ignoring the fact that it will break ls1012a without this being applied too.
I don't want to blame anyone, but I want to voice my disappointment and raise attention that something is obviously wrong here.
Hi Priyanka,
Could you please help in reviewing this patch. The idea/motivation of these changes are taken from fm driver. The said changes in fm are already accepted in commit 382c53f94631 ("net: fm: add TFABOOT support"). You may take a look at code here[1], line: 380.
Thanks Kuldeep [1] https://gitlab.denx.de/u-boot/u-boot/-/blob/master/drivers/net/fm/fm.c
-----Original Message----- From: Kuldeep Singh kuldeep.singh@nxp.com Sent: Friday, June 19, 2020 3:40 PM To: Joe Hershberger joe.hershberger@ni.com; u-boot@lists.denx.de; Priyanka Jain priyanka.jain@nxp.com Cc: Tom Rini trini@konsulko.com; Schrempf Frieder frieder.schrempf@kontron.de Subject: RE: [EXT] Re: [RESEND][Patch v4] net: pfe_eth: Use spi_flash_read API to access flash memory
-----Original Message----- From: Schrempf Frieder frieder.schrempf@kontron.de Sent: Thursday, May 28, 2020 1:46 PM To: Kuldeep Singh kuldeep.singh@nxp.com; Joe Hershberger joe.hershberger@ni.com; u-boot@lists.denx.de Cc: Priyanka Jain priyanka.jain@nxp.com; Tom Rini trini@konsulko.com Subject: [EXT] Re: [RESEND][Patch v4] net: pfe_eth: Use spi_flash_read API to access flash memory
Caution: EXT Email
On 28.05.20 08:12, Kuldeep Singh wrote:
Current PFE firmware access spi-nor memory directly. New spi-mem framework does not support direct memory access. So, let's use spi_flash_read API to access memory instead of directly using it.
Signed-off-by: Kuldeep Singh kuldeep.singh@nxp.com Reviewed-by: Frieder Schrempf frieder.schrempf@kontron.de
So this patch has been floating around for about half a year now with almost no attention from the maintainers! Several pings have been sent by the author without response.
In fact this patch was blocking 91afd36f3802 ("spi: Transform the FSL QuadSPI driver to use the SPI MEM API") that has finally been merged, ignoring the fact that it will break ls1012a without this being applied too.
I don't want to blame anyone, but I want to voice my disappointment and raise attention that something is obviously wrong here.
Hi Priyanka,
Could you please help in reviewing this patch. The idea/motivation of these changes are taken from fm driver. The said changes in fm are already accepted in commit 382c53f94631 ("net: fm: add TFABOOT support"). You may take a look at code here[1], line: 380.
Thanks Kuldeep [1] https://gitlab.denx.de/u-boot/u-boot/-/blob/master/drivers/net/fm/fm.c
Reviewed-by: Priyanka Jain priyanka.jain@nxp.com
Joe,
Kindly help to pick this patch.
Regards Priyanka
On Fri, Jun 19, 2020 at 02:13:37PM +0000, Priyanka Jain wrote:
-----Original Message----- From: Kuldeep Singh kuldeep.singh@nxp.com Sent: Friday, June 19, 2020 3:40 PM To: Joe Hershberger joe.hershberger@ni.com; u-boot@lists.denx.de; Priyanka Jain priyanka.jain@nxp.com Cc: Tom Rini trini@konsulko.com; Schrempf Frieder frieder.schrempf@kontron.de Subject: RE: [EXT] Re: [RESEND][Patch v4] net: pfe_eth: Use spi_flash_read API to access flash memory
-----Original Message----- From: Schrempf Frieder frieder.schrempf@kontron.de Sent: Thursday, May 28, 2020 1:46 PM To: Kuldeep Singh kuldeep.singh@nxp.com; Joe Hershberger joe.hershberger@ni.com; u-boot@lists.denx.de Cc: Priyanka Jain priyanka.jain@nxp.com; Tom Rini trini@konsulko.com Subject: [EXT] Re: [RESEND][Patch v4] net: pfe_eth: Use spi_flash_read API to access flash memory
Caution: EXT Email
On 28.05.20 08:12, Kuldeep Singh wrote:
Current PFE firmware access spi-nor memory directly. New spi-mem framework does not support direct memory access. So, let's use spi_flash_read API to access memory instead of directly using it.
Signed-off-by: Kuldeep Singh kuldeep.singh@nxp.com Reviewed-by: Frieder Schrempf frieder.schrempf@kontron.de
So this patch has been floating around for about half a year now with almost no attention from the maintainers! Several pings have been sent by the author without response.
In fact this patch was blocking 91afd36f3802 ("spi: Transform the FSL QuadSPI driver to use the SPI MEM API") that has finally been merged, ignoring the fact that it will break ls1012a without this being applied too.
I don't want to blame anyone, but I want to voice my disappointment and raise attention that something is obviously wrong here.
Hi Priyanka,
Could you please help in reviewing this patch. The idea/motivation of these changes are taken from fm driver. The said changes in fm are already accepted in commit 382c53f94631 ("net: fm: add TFABOOT support"). You may take a look at code here[1], line: 380.
Thanks Kuldeep [1] https://gitlab.denx.de/u-boot/u-boot/-/blob/master/drivers/net/fm/fm.c
Reviewed-by: Priyanka Jain priyanka.jain@nxp.com
Joe,
Kindly help to pick this patch.
This is something that should go via the NXP trees, thanks!
-----Original Message----- From: Tom Rini trini@konsulko.com Sent: Friday, June 19, 2020 8:07 PM To: Priyanka Jain priyanka.jain@nxp.com Cc: Kuldeep Singh kuldeep.singh@nxp.com; Joe Hershberger joe.hershberger@ni.com; u-boot@lists.denx.de; Schrempf Frieder frieder.schrempf@kontron.de Subject: Re: [EXT] Re: [RESEND][Patch v4] net: pfe_eth: Use spi_flash_read API to access flash memory
On Fri, Jun 19, 2020 at 02:13:37PM +0000, Priyanka Jain wrote:
-----Original Message----- From: Kuldeep Singh kuldeep.singh@nxp.com Sent: Friday, June 19, 2020 3:40 PM To: Joe Hershberger joe.hershberger@ni.com; u-boot@lists.denx.de; Priyanka Jain priyanka.jain@nxp.com Cc: Tom Rini trini@konsulko.com; Schrempf Frieder frieder.schrempf@kontron.de Subject: RE: [EXT] Re: [RESEND][Patch v4] net: pfe_eth: Use spi_flash_read API to access flash memory
-----Original Message----- From: Schrempf Frieder frieder.schrempf@kontron.de Sent: Thursday, May 28, 2020 1:46 PM To: Kuldeep Singh kuldeep.singh@nxp.com; Joe Hershberger joe.hershberger@ni.com; u-boot@lists.denx.de Cc: Priyanka Jain priyanka.jain@nxp.com; Tom Rini trini@konsulko.com Subject: [EXT] Re: [RESEND][Patch v4] net: pfe_eth: Use spi_flash_read API to access flash memory
Caution: EXT Email
On 28.05.20 08:12, Kuldeep Singh wrote:
Current PFE firmware access spi-nor memory directly. New spi-mem framework does not support direct memory access. So, let's use spi_flash_read API to access memory instead of directly using it.
Signed-off-by: Kuldeep Singh kuldeep.singh@nxp.com Reviewed-by: Frieder Schrempf frieder.schrempf@kontron.de
So this patch has been floating around for about half a year now with almost no attention from the maintainers! Several pings have been sent by the author without response.
In fact this patch was blocking 91afd36f3802 ("spi: Transform the FSL QuadSPI driver to use the SPI MEM API") that has finally been merged, ignoring the fact that it will break ls1012a without this being
applied too.
I don't want to blame anyone, but I want to voice my disappointment and raise attention that something is obviously wrong here.
Hi Priyanka,
Could you please help in reviewing this patch. The idea/motivation of these changes are taken from fm driver. The said changes in fm are already accepted in commit 382c53f94631
("net:
fm: add TFABOOT support"). You may take a look at code here[1], line: 380.
Thanks Kuldeep [1] https://gitlab.denx.de/u-boot/u-boot/-/blob/master/drivers/net/fm/fm. c
Reviewed-by: Priyanka Jain priyanka.jain@nxp.com
Joe,
Kindly help to pick this patch.
This is something that should go via the NXP trees, thanks!
Thanks Tom for mentioning. If this is so Priyanka, could you please pick this up in next PR.
-----Original Message----- From: Kuldeep Singh kuldeep.singh@nxp.com Sent: Monday, June 22, 2020 11:00 AM To: Priyanka Jain priyanka.jain@nxp.com Cc: Joe Hershberger joe.hershberger@ni.com; u-boot@lists.denx.de; Schrempf Frieder frieder.schrempf@kontron.de; Tom Rini trini@konsulko.com Subject: RE: [EXT] Re: [RESEND][Patch v4] net: pfe_eth: Use spi_flash_read API to access flash memory
-----Original Message----- From: Tom Rini trini@konsulko.com Sent: Friday, June 19, 2020 8:07 PM To: Priyanka Jain priyanka.jain@nxp.com Cc: Kuldeep Singh kuldeep.singh@nxp.com; Joe Hershberger joe.hershberger@ni.com; u-boot@lists.denx.de; Schrempf Frieder frieder.schrempf@kontron.de Subject: Re: [EXT] Re: [RESEND][Patch v4] net: pfe_eth: Use spi_flash_read API to access flash memory
On Fri, Jun 19, 2020 at 02:13:37PM +0000, Priyanka Jain wrote:
-----Original Message----- From: Kuldeep Singh kuldeep.singh@nxp.com Sent: Friday, June 19, 2020 3:40 PM To: Joe Hershberger joe.hershberger@ni.com; u-boot@lists.denx.de; Priyanka Jain priyanka.jain@nxp.com Cc: Tom Rini trini@konsulko.com; Schrempf Frieder frieder.schrempf@kontron.de Subject: RE: [EXT] Re: [RESEND][Patch v4] net: pfe_eth: Use spi_flash_read API to access flash memory
-----Original Message----- From: Schrempf Frieder frieder.schrempf@kontron.de Sent: Thursday, May 28, 2020 1:46 PM To: Kuldeep Singh kuldeep.singh@nxp.com; Joe Hershberger joe.hershberger@ni.com; u-boot@lists.denx.de Cc: Priyanka Jain priyanka.jain@nxp.com; Tom Rini trini@konsulko.com Subject: [EXT] Re: [RESEND][Patch v4] net: pfe_eth: Use spi_flash_read API to access flash memory
Caution: EXT Email
On 28.05.20 08:12, Kuldeep Singh wrote:
Current PFE firmware access spi-nor memory directly. New spi-mem framework does not support direct memory access. So, let's use spi_flash_read API to access memory instead of directly using
it.
Signed-off-by: Kuldeep Singh kuldeep.singh@nxp.com Reviewed-by: Frieder Schrempf frieder.schrempf@kontron.de
So this patch has been floating around for about half a year now with almost no attention from the maintainers! Several pings have been sent by the author without response.
In fact this patch was blocking 91afd36f3802 ("spi: Transform the FSL QuadSPI driver to use the SPI MEM API") that has finally been merged, ignoring the fact that it will break ls1012a without this being
applied too.
I don't want to blame anyone, but I want to voice my disappointment and raise attention that something is obviously wrong
here.
Hi Priyanka,
Could you please help in reviewing this patch. The idea/motivation of these changes are taken from fm driver. The said changes in fm are already accepted in commit 382c53f94631
("net:
fm: add TFABOOT support"). You may take a look at code here[1], line: 380.
Thanks Kuldeep [1] https://gitlab.denx.de/u-boot/u-boot/-/blob/master/drivers/net/fm/fm. c
Reviewed-by: Priyanka Jain priyanka.jain@nxp.com
Joe,
Kindly help to pick this patch.
This is something that should go via the NXP trees, thanks!
Thanks Tom for mentioning. If this is so Priyanka, could you please pick this up in next PR.
I will take care of this.
Thanks Priyanka
participants (4)
-
Kuldeep Singh -
Priyanka Jain -
Schrempf Frieder -
Tom Rini