Re: Weird MMC errors: 2 of 2 - inconsistent state after data crc error

Re: Weird MMC errors: 2 of 2 - inconsistent state after data crc error

[Alex Dubov]

Problem 2: After a data crc error all subsequent commands fail. May it be caused by stop command
leaving card in some bad state (something clearable by SEND_STATUS)? On the other hand, is there a
real need to issue a stop command in case main command failed?

---- Example: ----
Mar 14 09:25:13 Tirion kernel:  mmcblk0:<7>mmc0: starting CMD18 arg 00000000 flags 00000035
Mar 14 09:25:13 Tirion kernel: tifm_sd tifm_sd0:1: setting dma for 8 blocks
Mar 14 09:25:13 Tirion kernel: tifm_sd tifm_sd0:1: executing opcode 0x12, arg: 0x0, mask: 0xb100
Mar 14 09:25:13 Tirion kernel: tifm_sd tifm_sd0:1: host event: host_status 1, flags 0
Mar 14 09:25:13 Tirion kernel: tifm_sd tifm_sd0:1: host event: host_status 40, flags 1
Mar 14 09:25:13 Tirion kernel: tifm_sd tifm_sd0:1: executing opcode 0xc, arg: 0x0, mask: 0x2900
Mar 14 09:25:13 Tirion kernel: tifm_sd tifm_sd0:1: host event: host_status 80, flags 9
Mar 14 09:25:13 Tirion kernel: mmc0: req done (CMD18): 0/2/1: 00000900 00000000 00000000 00000000
Mar 14 09:25:13 Tirion kernel: mmcblk0: error 2 transferring data
Mar 14 09:25:13 Tirion kernel: end_request: I/O error, dev mmcblk0, sector 0
Mar 14 09:25:13 Tirion kernel: printk: 342 messages suppressed.
Mar 14 09:25:13 Tirion kernel: Buffer I/O error on device mmcblk0, logical block 0
Mar 14 09:25:13 Tirion kernel: mmc0: starting CMD18 arg 00000000 flags 00000035
Mar 14 09:25:13 Tirion kernel: tifm_sd tifm_sd0:1: setting dma for 8 blocks
Mar 14 09:25:13 Tirion kernel: tifm_sd tifm_sd0:1: executing opcode 0x12, arg: 0x0, mask: 0xb100
Mar 14 09:25:13 Tirion kernel: tifm_sd tifm_sd0:1: host event: host_status 80, flags 0
Mar 14 09:25:13 Tirion kernel: tifm_sd tifm_sd0:1: executing opcode 0xc, arg: 0x0, mask: 0x2900
Mar 14 09:25:13 Tirion kernel: tifm_sd tifm_sd0:1: host event: host_status 80, flags 8
Mar 14 09:25:13 Tirion kernel: mmc0: req done (CMD18): 1/0/1: 00000000 00000000 00000000 00000000



 
____________________________________________________________________________________
Sucker-punch spam with award-winning protection. 
Try the free Yahoo! Mail Beta.
http://advision.webevents.yahoo.com/mailbeta/features_spam.html

Re: Weird MMC errors: 2 of 2 - inconsistent state after data crc error

[Pierre Ossman]

Alex Dubov wrote:
> Problem 2: After a data crc error all subsequent commands fail. May it be caused by stop command
> leaving card in some bad state (something clearable by SEND_STATUS)? On the other hand, is there a
> real need to issue a stop command in case main command failed?
> 

It might be, depending on what the problem is. E.g. timeout might still mean the
card processed the command and will start sending data.

Anyway, CRC errors should be extremely rare so I'd guess that either the card or
the controller has gotten confused. In many cases the card will shut down when
it gets annoyed, so that might be what you're seeing here.

Other than that, I'm not sure I can help that much. The stop commands should
never wedge the card, so that isn't the issue (unless the card is buggy).

Rgds
-- 
     -- Pierre Ossman

  Linux kernel, MMC maintainer        http://www.kernel.org
  PulseAudio, core developer          http://pulseaudio.org
  rdesktop, core developer          http://www.rdesktop.org