linux-kernel.vger.kernel.org archive mirror
 help / color / mirror / Atom feed
* [PATCH] spi: document broadcom qspi driver as broken
@ 2017-07-21 22:00 Arnd Bergmann
  2017-07-25 21:39 ` Kamal Dasu
  0 siblings, 1 reply; 5+ messages in thread
From: Arnd Bergmann @ 2017-07-21 22:00 UTC (permalink / raw)
  To: Mark Brown
  Cc: Arnd Bergmann, Cyrille Pitchen, Geert Uytterhoeven,
	Hauke Mehrtens, Kamal Dasu, linux-spi, linux-kernel

The newly broadcom qspi driver in drivers/spi produces a build
warning when CONFIG_MTD is disabled:

include/linux/mtd/cfi.h:76:2: #warning No CONFIG_MTD_CFI_Ix selected. No NOR chip support can work. [-Werror=cpp]

I had suggested a workaround earlier, but Cyrille Pitchen explained
that actually the broadcom qspi is broken here and needs to be
fixed, see the lenghthy reply in patchwork.

As nothing has happened on that driver, this tries to at least
avoid the build failure, by marking the driver as broken unless
CONFIG_MTD is enabled. Also, I add a WARN_ON_ONCE runtime
that triggers when the spi-nor framework and the driver disagree
about the command opcode, which was one of several issues that
Cyrille pointed out.

My patch does not attempt to fix any of the actual bugs though,
it just tries to avoid the build error while highlighting the
problems. Ideally, someone would step up to create a tested
fix. If that doesn't happen, please merge my version instead
as a workaround.

Fixes: fa236a7ef240 ("spi: bcm-qspi: Add Broadcom MSPI driver")
Link: https://patchwork.kernel.org/patch/9334097/
Link: https://patchwork.kernel.org/patch/9624585/
Signed-off-by: Arnd Bergmann <arnd@arndb.de>
Cc: Cyrille Pitchen <cyrille.pitchen@atmel.com>
---
---
 drivers/spi/Kconfig        |  1 +
 drivers/spi/spi-bcm-qspi.c | 23 ++++++++++++-----------
 2 files changed, 13 insertions(+), 11 deletions(-)

diff --git a/drivers/spi/Kconfig b/drivers/spi/Kconfig
index 9b31351fe429..c7a80f9d6dd0 100644
--- a/drivers/spi/Kconfig
+++ b/drivers/spi/Kconfig
@@ -164,6 +164,7 @@ config SPI_BCM_QSPI
 	tristate "Broadcom BSPI and MSPI controller support"
 	depends on ARCH_BRCMSTB || ARCH_BCM || ARCH_BCM_IPROC || \
 			BMIPS_GENERIC || COMPILE_TEST
+	depends on BROKEN || MTD
 	default ARCH_BCM_IPROC
 	help
 	  Enables support for the Broadcom SPI flash and MSPI controller.
diff --git a/drivers/spi/spi-bcm-qspi.c b/drivers/spi/spi-bcm-qspi.c
index b19722ba908c..a388a3873552 100644
--- a/drivers/spi/spi-bcm-qspi.c
+++ b/drivers/spi/spi-bcm-qspi.c
@@ -349,12 +349,13 @@ static void bcm_qspi_bspi_set_xfer_params(struct bcm_qspi *qspi, u8 cmd_byte,
 	bcm_qspi_write(qspi, BSPI, BSPI_FLEX_MODE_ENABLE, flex_mode);
 }
 
-static int bcm_qspi_bspi_set_flex_mode(struct bcm_qspi *qspi, int width,
-				       int addrlen, int hp)
+static int bcm_qspi_bspi_set_flex_mode(struct bcm_qspi *qspi,
+				       struct spi_flash_read_message *msg,
+				       int width, int addrlen, int hp)
 {
 	int bpc = 0, bpp = 0;
 	u8 command = SPINOR_OP_READ_FAST;
-	int flex_mode = 1, rv = 0;
+	int flex_mode = 1;
 	bool spans_4byte = false;
 
 	dev_dbg(&qspi->pdev->dev, "set flex mode w %x addrlen %x hp %d\n",
@@ -405,15 +406,14 @@ static int bcm_qspi_bspi_set_flex_mode(struct bcm_qspi *qspi, int width,
 		}
 		break;
 	default:
-		rv = -EINVAL;
-		break;
+		return -EINVAL;
 	}
 
-	if (rv == 0)
-		bcm_qspi_bspi_set_xfer_params(qspi, command, bpp, bpc,
-					      flex_mode);
+	WARN_ON_ONCE(command != msg->read_opcode);
 
-	return rv;
+	bcm_qspi_bspi_set_xfer_params(qspi, command, bpp, bpc,
+					      flex_mode);
+	return 0;
 }
 
 static int bcm_qspi_bspi_set_override(struct bcm_qspi *qspi, int width,
@@ -461,6 +461,7 @@ static int bcm_qspi_bspi_set_override(struct bcm_qspi *qspi, int width,
 }
 
 static int bcm_qspi_bspi_set_mode(struct bcm_qspi *qspi,
+				  struct spi_flash_read_message *msg,
 				  int width, int addrlen, int hp)
 {
 	int error = 0;
@@ -491,7 +492,7 @@ static int bcm_qspi_bspi_set_mode(struct bcm_qspi *qspi,
 	}
 
 	if (qspi->xfer_mode.flex_mode)
-		error = bcm_qspi_bspi_set_flex_mode(qspi, width, addrlen, hp);
+		error = bcm_qspi_bspi_set_flex_mode(qspi, msg, width, addrlen, hp);
 
 	if (error) {
 		dev_warn(&qspi->pdev->dev,
@@ -1012,7 +1013,7 @@ static int bcm_qspi_flash_read(struct spi_device *spi,
 
 	io_width = msg->data_nbits ? msg->data_nbits : SPI_NBITS_SINGLE;
 	addrlen = msg->addr_width;
-	ret = bcm_qspi_bspi_set_mode(qspi, io_width, addrlen, -1);
+	ret = bcm_qspi_bspi_set_mode(qspi, msg, io_width, addrlen, -1);
 
 	if (!ret)
 		ret = bcm_qspi_bspi_flash_read(spi, msg);
-- 
2.9.0

^ permalink raw reply related	[flat|nested] 5+ messages in thread

* Re: [PATCH] spi: document broadcom qspi driver as broken
  2017-07-21 22:00 [PATCH] spi: document broadcom qspi driver as broken Arnd Bergmann
@ 2017-07-25 21:39 ` Kamal Dasu
  2017-07-26  7:03   ` Arnd Bergmann
  2017-07-26 16:24   ` Mark Brown
  0 siblings, 2 replies; 5+ messages in thread
From: Kamal Dasu @ 2017-07-25 21:39 UTC (permalink / raw)
  To: Arnd Bergmann
  Cc: Mark Brown, Cyrille Pitchen, Geert Uytterhoeven, Hauke Mehrtens,
	linux-spi, linux-kernel

Arnd, Cyrille,

I am working on fixing spi-bcm-qspi.c as per Cyrill's suggestion as
mentioned here : https://patchwork.kernel.org/patch/9624585/.
 And remove the use of SPINOR_OP_READ* and there by remove need to
include spi-nor.h.

Thanks
Kamal

On Fri, Jul 21, 2017 at 6:00 PM, Arnd Bergmann <arnd@arndb.de> wrote:
> The newly broadcom qspi driver in drivers/spi produces a build
> warning when CONFIG_MTD is disabled:
>
> include/linux/mtd/cfi.h:76:2: #warning No CONFIG_MTD_CFI_Ix selected. No NOR chip support can work. [-Werror=cpp]
>
> I had suggested a workaround earlier, but Cyrille Pitchen explained
> that actually the broadcom qspi is broken here and needs to be
> fixed, see the lenghthy reply in patchwork.
>
> As nothing has happened on that driver, this tries to at least
> avoid the build failure, by marking the driver as broken unless
> CONFIG_MTD is enabled. Also, I add a WARN_ON_ONCE runtime
> that triggers when the spi-nor framework and the driver disagree
> about the command opcode, which was one of several issues that
> Cyrille pointed out.
>
> My patch does not attempt to fix any of the actual bugs though,
> it just tries to avoid the build error while highlighting the
> problems. Ideally, someone would step up to create a tested
> fix. If that doesn't happen, please merge my version instead
> as a workaround.
>
> Fixes: fa236a7ef240 ("spi: bcm-qspi: Add Broadcom MSPI driver")
> Link: https://patchwork.kernel.org/patch/9334097/
> Link: https://patchwork.kernel.org/patch/9624585/
> Signed-off-by: Arnd Bergmann <arnd@arndb.de>
> Cc: Cyrille Pitchen <cyrille.pitchen@atmel.com>
> ---
> ---
>  drivers/spi/Kconfig        |  1 +
>  drivers/spi/spi-bcm-qspi.c | 23 ++++++++++++-----------
>  2 files changed, 13 insertions(+), 11 deletions(-)
>
> diff --git a/drivers/spi/Kconfig b/drivers/spi/Kconfig
> index 9b31351fe429..c7a80f9d6dd0 100644
> --- a/drivers/spi/Kconfig
> +++ b/drivers/spi/Kconfig
> @@ -164,6 +164,7 @@ config SPI_BCM_QSPI
>         tristate "Broadcom BSPI and MSPI controller support"
>         depends on ARCH_BRCMSTB || ARCH_BCM || ARCH_BCM_IPROC || \
>                         BMIPS_GENERIC || COMPILE_TEST
> +       depends on BROKEN || MTD
>         default ARCH_BCM_IPROC
>         help
>           Enables support for the Broadcom SPI flash and MSPI controller.
> diff --git a/drivers/spi/spi-bcm-qspi.c b/drivers/spi/spi-bcm-qspi.c
> index b19722ba908c..a388a3873552 100644
> --- a/drivers/spi/spi-bcm-qspi.c
> +++ b/drivers/spi/spi-bcm-qspi.c
> @@ -349,12 +349,13 @@ static void bcm_qspi_bspi_set_xfer_params(struct bcm_qspi *qspi, u8 cmd_byte,
>         bcm_qspi_write(qspi, BSPI, BSPI_FLEX_MODE_ENABLE, flex_mode);
>  }
>
> -static int bcm_qspi_bspi_set_flex_mode(struct bcm_qspi *qspi, int width,
> -                                      int addrlen, int hp)
> +static int bcm_qspi_bspi_set_flex_mode(struct bcm_qspi *qspi,
> +                                      struct spi_flash_read_message *msg,
> +                                      int width, int addrlen, int hp)
>  {
>         int bpc = 0, bpp = 0;
>         u8 command = SPINOR_OP_READ_FAST;
> -       int flex_mode = 1, rv = 0;
> +       int flex_mode = 1;
>         bool spans_4byte = false;
>
>         dev_dbg(&qspi->pdev->dev, "set flex mode w %x addrlen %x hp %d\n",
> @@ -405,15 +406,14 @@ static int bcm_qspi_bspi_set_flex_mode(struct bcm_qspi *qspi, int width,
>                 }
>                 break;
>         default:
> -               rv = -EINVAL;
> -               break;
> +               return -EINVAL;
>         }
>
> -       if (rv == 0)
> -               bcm_qspi_bspi_set_xfer_params(qspi, command, bpp, bpc,
> -                                             flex_mode);
> +       WARN_ON_ONCE(command != msg->read_opcode);
>
> -       return rv;
> +       bcm_qspi_bspi_set_xfer_params(qspi, command, bpp, bpc,
> +                                             flex_mode);
> +       return 0;
>  }
>
>  static int bcm_qspi_bspi_set_override(struct bcm_qspi *qspi, int width,
> @@ -461,6 +461,7 @@ static int bcm_qspi_bspi_set_override(struct bcm_qspi *qspi, int width,
>  }
>
>  static int bcm_qspi_bspi_set_mode(struct bcm_qspi *qspi,
> +                                 struct spi_flash_read_message *msg,
>                                   int width, int addrlen, int hp)
>  {
>         int error = 0;
> @@ -491,7 +492,7 @@ static int bcm_qspi_bspi_set_mode(struct bcm_qspi *qspi,
>         }
>
>         if (qspi->xfer_mode.flex_mode)
> -               error = bcm_qspi_bspi_set_flex_mode(qspi, width, addrlen, hp);
> +               error = bcm_qspi_bspi_set_flex_mode(qspi, msg, width, addrlen, hp);
>
>         if (error) {
>                 dev_warn(&qspi->pdev->dev,
> @@ -1012,7 +1013,7 @@ static int bcm_qspi_flash_read(struct spi_device *spi,
>
>         io_width = msg->data_nbits ? msg->data_nbits : SPI_NBITS_SINGLE;
>         addrlen = msg->addr_width;
> -       ret = bcm_qspi_bspi_set_mode(qspi, io_width, addrlen, -1);
> +       ret = bcm_qspi_bspi_set_mode(qspi, msg, io_width, addrlen, -1);
>
>         if (!ret)
>                 ret = bcm_qspi_bspi_flash_read(spi, msg);
> --
> 2.9.0
>

^ permalink raw reply	[flat|nested] 5+ messages in thread

* Re: [PATCH] spi: document broadcom qspi driver as broken
  2017-07-25 21:39 ` Kamal Dasu
@ 2017-07-26  7:03   ` Arnd Bergmann
  2017-07-26 19:43     ` Kamal Dasu
  2017-07-26 16:24   ` Mark Brown
  1 sibling, 1 reply; 5+ messages in thread
From: Arnd Bergmann @ 2017-07-26  7:03 UTC (permalink / raw)
  To: Kamal Dasu
  Cc: Mark Brown, Cyrille Pitchen, Geert Uytterhoeven, Hauke Mehrtens,
	linux-spi, Linux Kernel Mailing List

On Tue, Jul 25, 2017 at 11:39 PM, Kamal Dasu <kdasu.kdev@gmail.com> wrote:
> Arnd, Cyrille,
>
> I am working on fixing spi-bcm-qspi.c as per Cyrill's suggestion as
> mentioned here : https://patchwork.kernel.org/patch/9624585/.
>  And remove the use of SPINOR_OP_READ* and there by remove need to
> include spi-nor.h.

Ok, thanks a lot!

Do you think your fix will be applicable to stable backports, or should we
add a simple dependency first for the stable kernels and remove that
dependency again with the proper fix?

       Arnd

^ permalink raw reply	[flat|nested] 5+ messages in thread

* Re: [PATCH] spi: document broadcom qspi driver as broken
  2017-07-25 21:39 ` Kamal Dasu
  2017-07-26  7:03   ` Arnd Bergmann
@ 2017-07-26 16:24   ` Mark Brown
  1 sibling, 0 replies; 5+ messages in thread
From: Mark Brown @ 2017-07-26 16:24 UTC (permalink / raw)
  To: Kamal Dasu
  Cc: Arnd Bergmann, Cyrille Pitchen, Geert Uytterhoeven,
	Hauke Mehrtens, linux-spi, linux-kernel

[-- Attachment #1: Type: text/plain, Size: 736 bytes --]

On Tue, Jul 25, 2017 at 05:39:57PM -0400, Kamal Dasu wrote:
> Arnd, Cyrille,
> 
> I am working on fixing spi-bcm-qspi.c as per Cyrill's suggestion as
> mentioned here : https://patchwork.kernel.org/patch/9624585/.
>  And remove the use of SPINOR_OP_READ* and there by remove need to
> include spi-nor.h.

That's great - any ETA?  I'm tempted to apply Arnd's patch for now to
help anyone who runs into problems with it, your fix could then revert
it when you've got a proper fix.

Please don't top post, reply in line with needed context.  This allows
readers to readily follow the flow of conversation and understand what
you are talking about and also helps ensure that everything in the
discussion is being addressed.

[-- Attachment #2: signature.asc --]
[-- Type: application/pgp-signature, Size: 488 bytes --]

^ permalink raw reply	[flat|nested] 5+ messages in thread

* Re: [PATCH] spi: document broadcom qspi driver as broken
  2017-07-26  7:03   ` Arnd Bergmann
@ 2017-07-26 19:43     ` Kamal Dasu
  0 siblings, 0 replies; 5+ messages in thread
From: Kamal Dasu @ 2017-07-26 19:43 UTC (permalink / raw)
  To: Arnd Bergmann
  Cc: Mark Brown, Cyrille Pitchen, Geert Uytterhoeven, Hauke Mehrtens,
	linux-spi, Linux Kernel Mailing List

On Wed, Jul 26, 2017 at 3:03 AM, Arnd Bergmann <arnd@arndb.de> wrote:
> On Tue, Jul 25, 2017 at 11:39 PM, Kamal Dasu <kdasu.kdev@gmail.com> wrote:
>> Arnd, Cyrille,
>>
>> I am working on fixing spi-bcm-qspi.c as per Cyrill's suggestion as
>> mentioned here : https://patchwork.kernel.org/patch/9624585/.
>>  And remove the use of SPINOR_OP_READ* and there by remove need to
>> include spi-nor.h.
>
> Ok, thanks a lot!
>
> Do you think your fix will be applicable to stable backports, or should we
> add a simple dependency first for the stable kernels and remove that
> dependency again with the proper fix?
>

The  fix would be applicable to stable backport, we don't need to add
a dependency.

>        Arnd

Kamal

^ permalink raw reply	[flat|nested] 5+ messages in thread

end of thread, other threads:[~2017-07-26 19:43 UTC | newest]

Thread overview: 5+ messages (download: mbox.gz / follow: Atom feed)
-- links below jump to the message on this page --
2017-07-21 22:00 [PATCH] spi: document broadcom qspi driver as broken Arnd Bergmann
2017-07-25 21:39 ` Kamal Dasu
2017-07-26  7:03   ` Arnd Bergmann
2017-07-26 19:43     ` Kamal Dasu
2017-07-26 16:24   ` Mark Brown

This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).