mmc_test : some results

PATCH New Nicolas Ferre 3 participants 5 messages Jun 6, 2008 – Jun 10, 2008

Subsystems (2)

MMC, SD and SDIO Arm/Microchip (AT91) SoC support

lore

Share & tools

lore

Nicolas Ferre

author

2008-06-06 15:31 UTC #0 lore raw

Hi Pierre,

Here are the results of a mmc_test run played on at91_mci after applying this patch series :
http://lkml.org/lkml/2008/5/30/201

I skipped some of the tests just to concentrate on those ones.

Can you tell me if the test is correct : are the errors reported for the xfer_size tests the normal behavior (I assume yes because a failure is simulated) ?

~ # echo 1 > /sys/class/mmc_host/mmc0/mmc0\:e624/test
mmc0: Starting tests of card mmc0:e624...
mmc0: Test case 1. Basic write (no data verification)...
mmc0: Result: OK
mmc0: Test case 2. Basic read (no data verification)...
mmc0: Result: OK
mmc0: Test case 3. Basic write (with data verification)...
mmc0: Result: OK
mmc0: Test case 4. Basic read (with data verification)...
mmc0: Result: OK
mmc0: Test case 5. Multi-block write...
mmc0: Result: OK
mmc0: Test case 6. Multi-block read...
mmc0: Result: OK
mmc0: Test case 7. Correct xfer_size at write (start failure)...
mmc0: Result: ERROR (-84)
mmc0: Test case 8. Correct xfer_size at read (start failure)...
mmc0: Result: ERROR (-110)
mmc0: Test case 9. Correct xfer_size at write (midway failure)...
mmc0: Result: ERROR (-84)
mmc0: Test case 10. Correct xfer_size at read (midway failure)...
mmc0: Result: ERROR (-110)
mmc0: Tests completed.
~ #

Thanks, regards,

Signature · 1 line

Nicolas Ferre

Gururaja Hebbar K R

2008-06-09 06:04 UTC #1 lore raw

Hi Nicolas,

Could you please be kind enough to share the test program so that we can test mmc driver on other platforms.

i have a customized arm9 versatile board with mmci controller and
would like to test the same

TIA,

Regards
Gururaja

-----Original Message-----
From: [email protected] [mailto:[email protected]] On Behalf Of Nicolas Ferre
Sent: Friday, June 06, 2008 8:31 AM
To: Pierre Ossman; ARM Linux Mailing List; Linux Kernel list
Subject: mmc_test : some results

Hi Pierre,

Here are the results of a mmc_test run played on at91_mci after applying this patch series :
http://lkml.org/lkml/2008/5/30/201

I skipped some of the tests just to concentrate on those ones.

Can you tell me if the test is correct : are the errors reported for the xfer_size tests the normal behavior (I assume yes because a failure is simulated) ?

Thanks, regards,
--
Nicolas Ferre

-------------------------------------------------------------------
List admin:
http://lists.arm.linux.org.uk/mailman/listinfo/linux-arm-kernel

FAQ:        http://www.arm.linux.org.uk/mailinglists/faq.php
Etiquette:  http://www.arm.linux.org.uk/mailinglists/etiquette.php

Nicolas Ferre

author

2008-06-09 07:55 UTC #2 lore raw

Gururaja Hebbar K R :

Quoted 4 lines

Hi Nicolas,
Could you please be kind enough to share the test program so that we can test mmc driver on other platforms.

mmc_test is included in the mainline kernel :
http://lkml.org/lkml/2008/5/17/45

I just selected the tests I wanted to use in the source code (in mmc_test_cases tab).

Cheers,

Signature · 1 line

Nicolas Ferre

Pierre Ossman

2008-06-09 11:09 UTC #3 lore raw

In reply to Nicolas Ferre

Quoted 11 lines · On Fri, 06 Jun 2008 17:31:15 +0200 Nicolas Ferre <[email protected]>…

Hi Pierre,
Here are the results of a mmc_test run played on at91_mci after applying this patch series :
http://lkml.org/lkml/2008/5/30/201
I skipped some of the tests just to concentrate on those ones.
Can you tell me if the test is correct : are the errors reported for the xfer_size tests the normal behavior (I assume yes because a failure is simulated) ?

No. You're supposed to get an OK there as well. Basically it tests that the driver doesn't report some unexpected error (-ETIMEDOUT is the desired one) or overreports the number of written bytes.

Quoted 2 lines

~ # echo 1 > /sys/class/mmc_host/mmc0/mmc0\:e624/test

FYI, I've added some changes that will make that invocation only run test 1. Do echo "" for all tests in the future.

Rgds
Pierre

Nicolas Ferre

author

2008-06-10 09:27 UTC #4 lore raw

PATCH 1/1 at91_mci: manage cmd error and data error independently

In at91_mci_completed_command() function, this patch distinguishes command error and data error. It reports it in the corresponding error field.

Signed-off-by Nicolas Ferre <[email protected]>

---
Related to this thread :

Pierre Ossman :

Quoted 9 lines

Quoted 5 lines · On Fri, 06 Jun 2008 17:31:15 +0200 Nicolas Ferre <[email protected]>…
Hi Pierre,
Here are the results of a mmc_test run played on at91_mci after applying this patch series :
http://lkml.org/lkml/2008/5/30/201

Quoted 9 lines

Quoted 4 lines
I skipped some of the tests just to concentrate on those ones.
Can you tell me if the test is correct : are the errors reported for the xfer_size tests the normal behavior (I assume yes because a failure is simulated) ?
No. You're supposed to get an OK there as well. Basically it tests that the driver doesn't report some unexpected error (-ETIMEDOUT is the desired one) or overreports the number of written bytes.

Ok, so here is the first correction of the driver reporting the proper values in read.

It goes on top of this series :
http://lkml.org/lkml/2008/5/30/201

drivers/mmc/host/at91_mci.c | 28 ++++++++++++++++++++--------

1 files changed, 20 insertions(+), 8 deletions(-)

drivers/mmc/host/at91_mci.c +0 -0

index 9948fe1..86a3c5b 100644
@@ -663,6 +664,7 @@ static void at91_mci_process_next(struct at91mci_host *host)

static void at91_mci_completed_command(struct at91mci_host *host, unsigned int status)

{
	struct mmc_command *cmd = host->cmd;

+ struct mmc_data *data = cmd->data;

	at91_mci_write(host, AT91_MCI_IDR, 0xffffffff & ~(AT91_MCI_SDIOIRQA | AT91_MCI_SDIOIRQB));

@@ -685,15 +687,25 @@ static void at91_mci_completed_command(struct at91mci_host *host, unsigned int s

			cmd->error = 0;
		}
		else {

- if (status & (AT91_MCI_RTOE | AT91_MCI_DTOE))
- cmd->error = -ETIMEDOUT;
- else if (status & (AT91_MCI_RCRCE | AT91_MCI_DCRCE))
- cmd->error = -EILSEQ;
- else
- cmd->error = -EIO;
+ if (status & (AT91_MCI_DTOE | AT91_MCI_DCRCE)) {
+ if (data) {
+ if (status & AT91_MCI_DTOE)
+ data->error = -ETIMEDOUT;
+ else if (status & AT91_MCI_DCRCE)
+ data->error = -EILSEQ;
+ }
+ } else {
+ if (status & AT91_MCI_RTOE)
+ cmd->error = -ETIMEDOUT;
+ else if (status & AT91_MCI_RCRCE)
+ cmd->error = -EILSEQ;
+ else
+ cmd->error = -EIO;
+ }

- pr_debug("Error detected and set to %d (cmd = %d, retries = %d)\n",
- cmd->error, cmd->opcode, cmd->retries);
+ pr_debug("Error detected and set to %d/%d (cmd = %d, retries = %d)\n",
+ cmd->error, data ? data->error : 0,

+				 cmd->opcode, cmd->retries);
		}
	}
	else

Signature · 1 line

1.5.3.7