SD Driver : FAT: FAT read failed (blocknr 8) in sd card error while removing sd card during writing

DISCUSSION MADHAV SINGHCHAUHAN 2 participants 2 messages Jul 13, 2010 MMC, SD and SDIO

Share & tools

MADHAV SINGHCHAUHAN

author

2010-07-13 06:07 UTC #0 lore raw

Hi Madhav,

I have change the card remove sequence as per latest kernel source 2.6.34 .... after changing this i face the same issue.... is there any way to check card detect status at block layer ? or any other way to avoid it ....

Quoted 1 line

Quoted 1 line
Can i see your changes made by you..paste your diff here....Also even though you change ur card detect sequence

still you will get these messages ,,but somewhat you can reduce these messages if ur card removal sequence is efficient.. By the way does your Host controller support SDHCI??What you use in card detect interrupt handler workqueue or tasklet?? After removing card .... when i execute any command on board then also
i got below error messages :

Quoted 1 line

Quoted 1 line
Once you remove your card and if you still execute any command in the mount point of the card

you will get these type of errors ...ignore them...

FAT: Directory bread(block 3898) failed
FAT: Directory bread(block 3899) failed
FAT: Directory bread(block 3900) failed
FAT: Directory bread(block 3901) failed
FAT: Directory bread(block 3902) failed

Quoted 139 lines · MADHAV SINGHCHAUHAN

I am using 2.6.30 linux kernel version ...
Quoted 1 line
Quoted 1 line
As i told you previously that you need to flush queue as soon as possible once you remove the card.
So once you remove card mmc_blk_remove() gets called which calls mmc_cleanup_queue() and it further intenally calls kthread_stop()(stops thread which process requests).So compare the latest kernel version wrt your kernel version and you will find the differences in card remove sequence so either you can backport these functions or change manually sequence of these functions.
Quoted 111 lines · Madhav Chauhan
Quoted 8 lines · hardik patel
Hi Madhav,
thanks for your interest... I have card detect register so mmc_get_cd function i can check whether card is inserted or not ... so in sd host controller driver can detect that properly... but i think block driver does not detect that .... where i have to flush my request queue... i have gone through block.c and queue.c file which generate mmc_block.ko file but not sure where to modify... can any one guide me more
Which Kernel version you are using??
Quoted 99 lines
Quoted 81 lines · MADHAV SINGHCHAUHAN
Card removal not detected by driver?
You should see a line of log if card removal is detected:
card xxxx removed
Would you describe how do you detect card removal/insertion on your board?
Regards,
-Ethan
Quoted 59 lines · hardik patel
I have sd card on board ... During recording/writing SD card, remove SD card ... then i am getting endless error in dmesg as below :
end_request: I/O error, dev mmcblk0, sector 24
FAT: FAT read failed (blocknr 8)
mmcblk0: error -110 sending status comand<3>mmcblk0: error -110 sending read/write command, response 0x0, card status 0x0 Block mmc_blk_issue_rq function call 346
end_request: I/O error, dev mmcblk0, sector 4
FAT: FAT read failed (blocknr 4)
MMC: killing requests for dead queue
end_request: I/O error, dev mmcblk0, sector 8
FAT: FAT read failed (blocknr 8)
FAT: FAT read failed (blocknr 4)
FAT: FAT read failed (blocknr 8)
FAT: FAT read failed (blocknr 4)
FAT: FAT read failed (blocknr 8)
FAT: FAT read failed (blocknr 4)
FAT: FAT read failed (blocknr 8)
How to stop this message ... what is ideal situation when we remove card while writing ...... can u help me what to do ?
Quoted 4 lines
When we insert the card hot plug dameon start sending the request to block layer and kernel block layer start sending the request to card.But when you remove the card in between request queue is not flushed properly.. so you need to flush your request queue at appropriate time.Also please note that even that you flush your request queue at appropriate time some messages will come.
where is the problem here? sd host controler driver problem or block driver problem or file system problem
Note: I have formatted sd card with FAT32 file system .
With EXT3 File system :
Remove SD card while writing/recording we got error messages like below. but this is not endless. After some time it stops.
==============================
=============
end_request: I/O error, dev mmcblk0, sector 156536
mmcblk0: error -110 sending status comand<3>mmcblk0: error -110 sending read/write command, response 0x0, card status 0x0
mmcblk0: error -110 requesting status
end_request: I/O error, dev mmcblk0, sector 4736 Aborting journal on device mmcblk0p1.
ext3_abort called.
EXT3-fs error (device mmcblk0p1): ext3_journal_start_sb: Detected aborted journal
Remounting filesystem read-only
end_request: I/O error, dev mmcblk0, sector 4448 journal commit I/O error
===========================================
Signature · 3 lines
To unsubscribe from this list: send the line "unsubscribe linux-mmc" in 
the body of a message to [email protected] 
More majordomo info at ?http://vger.kernel.org/majordomo-info.html 
Signature · 3 lines
To unsubscribe from this list: send the line "unsubscribe linux-mmc" in 
the body of a message to [email protected] 
More majordomo info at  http://vger.kernel.org/majordomo-info.html 
Signature · 9 lines
Regards, 
Hardik 

"If you are doing what you have always been doing then expect what you 
have always been getting." 
-- 
To unsubscribe from this list: send the line "unsubscribe linux-mmc" in 
the body of a message to [email protected] 
More majordomo info at  http://vger.kernel.org/majordomo-info.html 
Signature · 5 lines
Regards, 
Hardik 

"If you are doing what you have always been doing then expect what you 
have always been getting." 

Signature · 5 lines

Regards, 
Hardik 

"If you are doing what you have always been doing then expect what you 
have always been getting."

hardik patel

2010-07-13 09:24 UTC #1 lore raw

Find below diff :

queue_2_6_34.c 2010-07-13 13:54:35.000000000 +0530 renamed from queue.c 2010-07-13 13:52:53.000000000 +0530 +6 -7

@@ -236,17 +236,18 @@ void mmc_cleanup_queue(struct mmc_queue
        struct request_queue *q = mq->queue;
        unsigned long flags;
 
-       /* Mark that we should start throwing out stragglers */
-       spin_lock_irqsave(q->queue_lock, flags);
-       q->queuedata = NULL;
-       spin_unlock_irqrestore(q->queue_lock, flags);
-
        /* Make sure the queue isn't suspended, as that will deadlock */
        mmc_queue_resume(mq);
 
        /* Then terminate our worker thread */
        kthread_stop(mq->thread);
 
+       /* Empty the queue */
+        spin_lock_irqsave(q->queue_lock, flags);
+        q->queuedata = NULL;
+        blk_start_queue(q);
+        spin_unlock_irqrestore(q->queue_lock, flags);
+
        if (mq->bounce_sg)
                kfree(mq->bounce_sg);
        mq->bounce_sg = NULL;

@@ -258,8 +259,6 @@ void mmc_cleanup_queue(struct mmc_queue
                kfree(mq->bounce_buf);
        mq->bounce_buf = NULL;
 
-       blk_cleanup_queue(mq->queue);
-
        mq->card = NULL;
 }
 EXPORT_SYMBOL(mmc_cleanup_queue);

In card detect interrupt handler we have use tasklet.... no my Host controller does not support SDHCI.

Quoted 178 lines · MADHAV SINGHCHAUHAN

Hi Madhav,
I have change the card remove sequence as per latest kernel source 2.6.34 .... after changing this i face the same issue.... is there any way to check card detect status at block layer ? or any other way to avoid it ....
Quoted 1 line
Quoted 1 line
Can i see your changes made by you..paste your diff here....Also even though you change ur card detect sequence
still you will get these messages ,,but somewhat you can reduce these messages if ur card removal sequence is efficient.. By the way does your Host controller support SDHCI??What you use in card detect interrupt handler workqueue or tasklet?? After removing card .... when i execute any command on board then also
i got below error messages :
Quoted 1 line
Quoted 1 line
Once you remove your card and if you still execute any command in the mount point of the card
you will get these type of errors ...ignore them...
FAT: Directory bread(block 3898) failed
FAT: Directory bread(block 3899) failed
FAT: Directory bread(block 3900) failed
FAT: Directory bread(block 3901) failed
FAT: Directory bread(block 3902) failed
Quoted 125 lines · MADHAV SINGHCHAUHAN
I am using 2.6.30 linux kernel version ...
Quoted 1 line
Quoted 1 line
As i told you previously that you need to flush queue as soon as possible once you remove the card.
So once you remove card mmc_blk_remove() gets called which calls mmc_cleanup_queue() and it further intenally calls kthread_stop()(stops thread which process requests).So compare the latest kernel version wrt your kernel version and you will find the differences in card remove sequence so either you can backport these functions or change manually sequence of these functions.
Quoted 111 lines · Madhav Chauhan
Quoted 8 lines · hardik patel
Hi Madhav,
thanks for your interest... I have card detect register so mmc_get_cd function i can check whether card is inserted or not ... so in sd host controller driver can detect that properly... but i think block driver does not detect that .... where i have to flush my request queue... i have gone through block.c and queue.c file which generate mmc_block.ko file but not sure where to modify... can any one guide me more
Which Kernel version you are using??
Quoted 99 lines
Quoted 69 lines · MADHAV SINGHCHAUHAN
Card removal not detected by driver?
You should see a line of log if card removal is detected:
card xxxx removed
Would you describe how do you detect card removal/insertion on your board?
Regards,
-Ethan
Quoted 50 lines · hardik patel
I have sd card on board ... During recording/writing SD card, remove SD card ... then i am getting endless error in dmesg as below :
end_request: I/O error, dev mmcblk0, sector 24
FAT: FAT read failed (blocknr 8)
mmcblk0: error -110 sending status comand<3>mmcblk0: error -110 sending read/write command, response 0x0, card status 0x0 Block mmc_blk_issue_rq function call 346
end_request: I/O error, dev mmcblk0, sector 4
FAT: FAT read failed (blocknr 4)
MMC: killing requests for dead queue
end_request: I/O error, dev mmcblk0, sector 8
FAT: FAT read failed (blocknr 8)
FAT: FAT read failed (blocknr 4)
FAT: FAT read failed (blocknr 8)
FAT: FAT read failed (blocknr 4)
FAT: FAT read failed (blocknr 8)
FAT: FAT read failed (blocknr 4)
FAT: FAT read failed (blocknr 8)
How to stop this message ... what is ideal situation when we remove card while writing ...... can u help me what to do ?
Quoted 4 lines
When we insert the card hot plug dameon start sending the request to block layer and kernel block layer start sending the request to card.But when you remove the card in between request queue is not flushed properly.. so you need to flush your request queue at appropriate time.Also please note that even that you flush your request queue at appropriate time some messages will come.
where is the problem here? sd host controler driver problem or block driver problem or file system problem
Note: I have formatted sd card with FAT32 file system .
With EXT3 File system :
Remove SD card while writing/recording we got error messages like below. but this is not endless. After some time it stops.
==============================
=============
end_request: I/O error, dev mmcblk0, sector 156536
mmcblk0: error -110 sending status comand<3>mmcblk0: error -110 sending read/write command, response 0x0, card status 0x0
mmcblk0: error -110 requesting status
end_request: I/O error, dev mmcblk0, sector 4736 Aborting journal on device mmcblk0p1.
ext3_abort called.
EXT3-fs error (device mmcblk0p1): ext3_journal_start_sb: Detected aborted journal
Quoted 9 lines
Remounting filesystem read-only
end_request: I/O error, dev mmcblk0, sector 4448
journal commit I/O error
===========================================
--
To unsubscribe from this list: send the line "unsubscribe linux-mmc" in the body of a message to [email protected]
More majordomo info at ?http://vger.kernel.org/majordomo-info.html
Quoted 12 lines
--
To unsubscribe from this list: send the line "unsubscribe linux-mmc" in the body of a message to [email protected]
More majordomo info at http://vger.kernel.org/majordomo-info.html
--
Regards,
Hardik
"If you are doing what you have always been doing then expect what you have always been getting."
--
To unsubscribe from this list: send the line "unsubscribe linux-mmc" in the body of a message to [email protected]
More majordomo info at http://vger.kernel.org/majordomo-info.html
Quoted 14 lines
--
Regards,
Hardik
"If you are doing what you have always been doing then expect what you have always been getting."
--
Regards,
Hardik
"If you are doing what you have always been doing then expect what you have always been getting."

Signature · 5 lines

Regards,
Hardik

"If you are doing what you have always been doing then expect what you
have always been getting."