All of lore.kernel.org
 help / color / mirror / Atom feed
* [PATCH] serial: 8250_exar: Bug fix for determination of number of ports for Commtech PCIe cards
@ 2020-08-12 21:49 Valmer Huhn
  2020-08-13 13:51 ` Andy Shevchenko
  0 siblings, 1 reply; 3+ messages in thread
From: Valmer Huhn @ 2020-08-12 21:49 UTC (permalink / raw)
  To: gregkh, andriy.shevchenko, sudip.mukherjee, jan.kiszka; +Cc: linux-serial

serial: 8250_exar: Bug fix for determination of number of ports for
Commtech PCIe cards

The following line is used to determine the number of ports for each exar
board in a/drivers/tty/serial/8250/8250_exar.c:589

nr_ports = board->num_ports ? board->num_ports : pcidev->device & 0x0f;

If the number of ports a card has is not explicitly specified, it defaults
to the rightmost 4 bits of the PCI device ID. This is prone to error since
not all PCI device IDs contain a number which corresponds to the number of
ports that card provides.

This particular case involves COMMTECH_4224PCIE and COMMTECH_4228PCIE
cards with device ID 0x0020 and 0x0021. Currently the multiport cards
receive 0 and 1 port instead of 4 and 8 ports respectively.

To fix this, each Commtech Fastcom PCIe card is given a struct where the
number of ports is explicitly specified. This ensures 'board->num_ports'
is used instead of the default 'pcidev->device & 0x0f'.

Signed-off-by: Valmer Huhn <valmer.huhn@concurrent-rt.com>
Tested-by: Valmer Huhn <valmer.huhn@concurrent-rt.com>

---
Thank you,
Valmer Huhn

diff --git a/drivers/tty/serial/8250/8250_exar.c b/drivers/tty/serial/8250/8250_exar.c
index ddb6aeb76dc5..793906afb0f6 100644
--- a/drivers/tty/serial/8250/8250_exar.c
+++ b/drivers/tty/serial/8250/8250_exar.c
@@ -734,6 +734,24 @@ static const struct exar8250_board pbn_exar_XR17V35x = {
 	.exit		= pci_xr17v35x_exit,
 };
 
+static const struct exar8250_board pbn_fastcom_XR17V352 = {
+	.num_ports	= 2,
+	.setup		= pci_xr17v35x_setup,
+	.exit		= pci_xr17v35x_exit,
+};
+
+static const struct exar8250_board pbn_fastcom_XR17V354 = {
+	.num_ports	= 4,
+	.setup		= pci_xr17v35x_setup,
+	.exit		= pci_xr17v35x_exit,
+};
+
+static const struct exar8250_board pbn_fastcom_XR17V358 = {
+	.num_ports	= 8,
+	.setup		= pci_xr17v35x_setup,
+	.exit		= pci_xr17v35x_exit,
+};
+
 static const struct exar8250_board pbn_exar_XR17V4358 = {
 	.num_ports	= 12,
 	.setup		= pci_xr17v35x_setup,
@@ -801,9 +819,9 @@ static const struct pci_device_id exar_pci_tbl[] = {
 	EXAR_DEVICE(EXAR, XR17V358, pbn_exar_XR17V35x),
 	EXAR_DEVICE(EXAR, XR17V4358, pbn_exar_XR17V4358),
 	EXAR_DEVICE(EXAR, XR17V8358, pbn_exar_XR17V8358),
-	EXAR_DEVICE(COMMTECH, 4222PCIE, pbn_exar_XR17V35x),
-	EXAR_DEVICE(COMMTECH, 4224PCIE, pbn_exar_XR17V35x),
-	EXAR_DEVICE(COMMTECH, 4228PCIE, pbn_exar_XR17V35x),
+	EXAR_DEVICE(COMMTECH, 4222PCIE, pbn_fastcom_XR17V352),
+	EXAR_DEVICE(COMMTECH, 4224PCIE, pbn_fastcom_XR17V354),
+	EXAR_DEVICE(COMMTECH, 4228PCIE, pbn_fastcom_XR17V358),
 
 	EXAR_DEVICE(COMMTECH, 4222PCI335, pbn_fastcom335_2),
 	EXAR_DEVICE(COMMTECH, 4224PCI335, pbn_fastcom335_4),

^ permalink raw reply related	[flat|nested] 3+ messages in thread

* Re: [PATCH] serial: 8250_exar: Bug fix for determination of number of ports for Commtech PCIe cards
  2020-08-12 21:49 [PATCH] serial: 8250_exar: Bug fix for determination of number of ports for Commtech PCIe cards Valmer Huhn
@ 2020-08-13 13:51 ` Andy Shevchenko
  2020-08-13 16:42   ` Valmer Huhn
  0 siblings, 1 reply; 3+ messages in thread
From: Andy Shevchenko @ 2020-08-13 13:51 UTC (permalink / raw)
  To: Valmer Huhn; +Cc: gregkh, sudip.mukherjee, jan.kiszka, linux-serial

On Wed, Aug 12, 2020 at 05:49:37PM -0400, Valmer Huhn wrote:

Thanks for the report. It needs a slight corrections I tell about below.

> serial: 8250_exar: Bug fix for determination of number of ports for
> Commtech PCIe cards

This should not be present in the body.
And in the subject it would be better to reduce the text to something like

serial: 8250_exar: Fix number of ports for Commtech PCIe cards

> The following line is used to determine the number of ports for each exar
> board in a/drivers/tty/serial/8250/8250_exar.c:589

There is no a/ folder and we don't need the full path anyway, just something
like '8250_exar.c line 589' is enough.

Also refer to Exar with capitalized name.

> nr_ports = board->num_ports ? board->num_ports : pcidev->device & 0x0f;
> 
> If the number of ports a card has is not explicitly specified, it defaults
> to the rightmost 4 bits of the PCI device ID. This is prone to error since
> not all PCI device IDs contain a number which corresponds to the number of
> ports that card provides.
> 
> This particular case involves COMMTECH_4224PCIE and COMMTECH_4228PCIE
> cards with device ID 0x0020 and 0x0021. Currently the multiport cards

'...with device IDs 0x0020, 0x0021 and 0x0022.'

> receive 0 and 1 port instead of 4 and 8 ports respectively.

and update this accordingly.

> To fix this, each Commtech Fastcom PCIe card is given a struct where the
> number of ports is explicitly specified. This ensures 'board->num_ports'
> is used instead of the default 'pcidev->device & 0x0f'.

Please, add a Fixes tag.

Fixes: d0aeaa83f0b0 ("serial: exar: split out the exar code from 8250_pci")

> Signed-off-by: Valmer Huhn <valmer.huhn@concurrent-rt.com>
> Tested-by: Valmer Huhn <valmer.huhn@concurrent-rt.com>

...

> -	EXAR_DEVICE(COMMTECH, 4222PCIE, pbn_exar_XR17V35x),
> -	EXAR_DEVICE(COMMTECH, 4224PCIE, pbn_exar_XR17V35x),
> -	EXAR_DEVICE(COMMTECH, 4228PCIE, pbn_exar_XR17V35x),
> +	EXAR_DEVICE(COMMTECH, 4222PCIE, pbn_fastcom_XR17V352),
> +	EXAR_DEVICE(COMMTECH, 4224PCIE, pbn_fastcom_XR17V354),
> +	EXAR_DEVICE(COMMTECH, 4228PCIE, pbn_fastcom_XR17V358),

For sake of the consistency I would rather see them as

pbn_fastcom35x_2/4/8.

>  	EXAR_DEVICE(COMMTECH, 4222PCI335, pbn_fastcom335_2),
>  	EXAR_DEVICE(COMMTECH, 4224PCI335, pbn_fastcom335_4),

-- 
With Best Regards,
Andy Shevchenko



^ permalink raw reply	[flat|nested] 3+ messages in thread

* Re: [PATCH] serial: 8250_exar: Bug fix for determination of number of ports for Commtech PCIe cards
  2020-08-13 13:51 ` Andy Shevchenko
@ 2020-08-13 16:42   ` Valmer Huhn
  0 siblings, 0 replies; 3+ messages in thread
From: Valmer Huhn @ 2020-08-13 16:42 UTC (permalink / raw)
  To: Andy Shevchenko; +Cc: gregkh, sudip.mukherjee, jan.kiszka, linux-serial

On Thu, Aug 13, 2020 at 04:51:11PM +0300, Andy Shevchenko wrote:
> On Wed, Aug 12, 2020 at 05:49:37PM -0400, Valmer Huhn wrote:
> 
> Thanks for the report. It needs a slight corrections I tell about below.

Thank you very much for your feedback.

> > serial: 8250_exar: Bug fix for determination of number of ports for
> > Commtech PCIe cards
> 
> This should not be present in the body.
> And in the subject it would be better to reduce the text to something like
> 
> serial: 8250_exar: Fix number of ports for Commtech PCIe cards

I will resubmit the patch with this subject line and include the changes
you suggested.

Thank you,
Valmer Huhn

^ permalink raw reply	[flat|nested] 3+ messages in thread

end of thread, other threads:[~2020-08-13 16:43 UTC | newest]

Thread overview: 3+ messages (download: mbox.gz / follow: Atom feed)
-- links below jump to the message on this page --
2020-08-12 21:49 [PATCH] serial: 8250_exar: Bug fix for determination of number of ports for Commtech PCIe cards Valmer Huhn
2020-08-13 13:51 ` Andy Shevchenko
2020-08-13 16:42   ` Valmer Huhn

This is an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.