Re: [BUG]: mmc: single thread workqueue very bad - breaks multiplePERMANENT CARDS

From: Philip Rakity
Date: Mon Dec 13 2010 - 19:03:25 EST

Next message: Stephen Rothwell: "Re: rcu-walk and dcache scaling tree update and status"
Previous message: Mike Travis: "Early kernel messages are overflowing the static log buffer"
In reply to: Jaehoon Chung: "Re: [BUG]: mmc: single thread workqueue very bad - breaks multiplePERMANENT CARDS"
Messages sorted by: [ date ] [ thread ] [ subject ] [ author ]

rootdelay

if the boot device is the 1st device the arch code installs it seems to work otherwise try
changing create_singlethread_workqueue to create_workqueue.

>>> The following change makes things work --- IS THIS SAFE -?
>>>
>>> static int __init mmc_init(void)
>>> {
>>> int ret;
>>>
>>> #if 0
>>> workqueue = create_singlethread_workqueue("kmmcd");
>>> #else
>>> workqueue = create_workqueue("kmmcd");
>>> #endif
>>> if (!workqueue)
>>> return -ENOMEM;
>>>
>>

On Dec 13, 2010, at 3:53 PM, Jaehoon Chung wrote:

> Hi Philip.
>
> I known that have kernel parameter.."rootwait"
>
> Did you use "rootwait" or "rootdelay"?
> I also need information about this problem..
>
>
> Philip Rakity wrote:
>> Needed to change in drivers/mmc/core
>>
>>> create_singlethread_workqueue
>> to
>> create_workqueue
>>
>> to make multiple cards be detected at boot time.
>>
>> Logs with traces (not working and working) below.
>>
>> This could be a bug in the mmc layer or in the workqueue code.
>>
>> reposting to linux-list and linux-mmc for additional comments
>>
>>>
>>> I am trying to boot
>>>
>>> I have an external SD card.
>>>
>>> I have a PERMANENT eMMC card on the board. (BROKEN_CARD_DETECT set and POLLING OFF)
>>> I have a PERMANENT sdio card on the board (BROKEN_CARD_DETECT set and POLLING OFF)
>>> I am trying to boot from the eMMC.
>>>
>>> in the arch directory if I init the SD controller first -- booting fails
>>> if I init the eMMC SD booting works
>>>
>>> From the trace so far it looks like I cannot have 2 entries on the the workqueue for mmc_rescan.
>>>
>>> The sdhci driver sees 2 permanent slots and calls up to the mmc layer to handle initial card detect.
>>>
>>> I think the workqueue needs to check if work is pending and reschedule itself -- not sure how to do this.
>>>
>>> mmc_rescan is not being called -- any ideas ?
>>>
>>> trace below
>>>
>>> max8925-rtc max8925-rtc: rtc core: registered max8925-rtc as rtc0
>>> [ 89.784231] sdhci: Secure Digital Host Controller Interface driver
>>> [ 89.802610] sdhci: Copyright(c) Pierre Ossman
>>> [ 89.815573] mmc0: no vmmc regulator found
>>> [ 89.827627] mmc_start_host:mmc0: ENTER
>>> [ 89.838892] mmc_detect_change: mmc0: ENTER
>>> [ 89.851084] mmc_schedule_delayed_work: ENTER
>>> [ 89.863893] mmc_schedule_delayed_work: x = 1 EXIT
>>> [ 89.877922] mmc_rescan:mmc0: ENTER
>>> [ 89.888025] mmc0: mmc_rescan: trying to init card at 400000 Hz
>>> [ 89.905442] mmc_detect_change: mmc0: EXIT
>>> [ 89.917422] mmc0: SDHCI controller on MMC [sdhci-pxa.0] using DMA
>>> [ 89.935552] mmc1: no vmmc regulator found
>>> [ 89.950739] NOT SDIO/SD/MMC
>>> [ 89.959061] mmc_rescan:mmc0 FAIL
>>> [ 89.968661] mmc0: mmc_rescan: trying to init card at 300000 Hz
>>> [ 89.986039] mmc_start_host:mmc1: ENTER
>>> [ 89.997404] mmc_detect_change: mmc1: ENTER
>>> [ 90.009585] mmc_schedule_delayed_work: ENTER
>>> [ 90.025342] mmc_schedule_delayed_work: x = 1 EXIT
>>> [ 90.039396] NOT SDIO/SD/MMC
>>> [ 90.047703] mmc_rescan:mmc0 FAIL
>>> [ 90.057304] mmc0: mmc_rescan: trying to init card at 200000 Hz
>>> [ 90.074664] mmc_detect_change: mmc1: EXIT
>>> [ 90.086698] mmc1: SDHCI controller on MMC [sdhci-pxa.1] using DMA
>>> [ 90.104817] mmc2: no vmmc regulator found
>>> [ 90.119917] mmc_start_host:mmc2: ENTER
>>> [ 90.131280] NOT SDIO/SD/MMC
>>> [ 90.139585] mmc_rescan:mmc0 FAIL
>>> [ 90.149271] mmc0: mmc_rescan: trying to init card at 100000 Hz
>>> [ 90.166643] mmc_detect_change: mmc2: ENTER
>>> [ 90.178818] mmc_schedule_delayed_work: ENTER
>>> [ 90.191703] mmc_schedule_delayed_work: x = 1 EXIT
>>> [ 90.205696] mmc_detect_change: mmc2: EXIT
>>> [ 90.220729] NOT SDIO/SD/MMC
>>> [ 90.229042] mmc_rescan:mmc0 FAIL
>>> [ 90.238638] mmc2: SDHCI controller on MMC [sdhci-pxa.2] using DMA
>>> [ 90.256770] TCP cubic registered
>>> [ 90.256954] NET: Registered protocol family 17
>>> [ 90.279805] Registering the dns_resolver key type
>>> [ 90.293787] VFP support v0.3: implementor 56 architecture 2 part 20 variant 9 rev 5
>>> [ 90.318180] max8925-rtc max8925-rtc: setting system clock to 2007-02-05 22:39:20 UTC (1170715160)
>>> [ 90.344586] Waiting 5sec before mounting root device...
>>> [ 95.366580] Root-NFS: no NFS server address
>>> [ 95.379055] VFS: Unable to mount root fs via NFS, trying floppy.
>>> [ 95.397037] mmc_rescan:mmc1: ENTER
>>> [ 95.407528] VFS: Cannot open root device "mmcblk0p1" or unknown-block(2,0)
>>> [ 95.428065] mmc1: mmc_rescan: trying to init card at 400000 Hz
>>> [ 95.445444] Please append a correct "root=" boot option; here are the available partitions:
>>> [ 95.470345] Kernel panic - not syncing: VFS: Unable to mount root fs on unknown-block(2,0)
>>> [ 95.494941] [<c002a9a8>] (unwind_backtrace+0x0/0x128) from [<c0274dc8>] (panic+0x70/0x184)
>>> [ 95.519606] [<c0274dc8>] (panic+0x70/0x184) from [<c0008d50>] (mount_block_root+0x1cc/0x20c)
>>> [ 95.544688] [<c0008d50>] (mount_block_root+0x1cc/0x20c) from [<c0008e30>] (mount_root+0xa0/0xc0)
>>> [ 95.570868] [<c0008e30>] (mount_root+0xa0/0xc0) from [<c0008f6c>] (prepare_namespace+0x11c/0x174)
>>> [ 95.597260] [<c0008f6c>] (prepare_namespace+0x11c/0x174) from [<c0008444>] (kernel_init+0x110/0x150)
>>> [ 95.624473] [<c0008444>] (kernel_init+0x110/0x150) from [<c00260e0>] (kernel_thread_exit+0x0/0x8)
>>>
>>>
>>>
>>> The following change makes things work --- IS THIS SAFE -?
>>>
>>> static int __init mmc_init(void)
>>> {
>>> int ret;
>>>
>>> #if 0
>>> workqueue = create_singlethread_workqueue("kmmcd");
>>> #else
>>> workqueue = create_workqueue("kmmcd");
>>> #endif
>>> if (!workqueue)
>>> return -ENOMEM;
>>>
>>>
>>> [ 1.680952] mousedev: PS/2 mouse device common for all mice
>>> [ 1.698753] max8925-rtc max8925-rtc: rtc core: registered max8925-rtc as rtc0
>>> [ 1.720238] sdhci: Secure Digital Host Controller Interface driver
>>> [ 1.738743] sdhci: Copyright(c) Pierre Ossman
>>> [ 1.751705] mmc0: no vmmc regulator found
>>> [ 1.763764] mmc_start_host:mmc0: ENTER
>>> [ 1.775028] mmc_detect_change: mmc0: work = df94bd50 ENTER
>>> [ 1.791367] mmc_schedule_delayed_work: work = df94bd50 ENTER
>>> [ 1.808194] mmc_rescan:mmc0: work = df94bd50 ENTER
>>> [ 1.822464] mmc_schedule_delayed_work: x = 1 EXIT
>>> [ 1.836436] mmc0: mmc_rescan: trying to init card at 400000 Hz
>>> [ 1.853812] mmc_detect_change: mmc0: EXIT
>>> [ 1.865717] mmc0: SDHCI controller on MMC [sdhci-pxa.0] using DMA
>>> [ 1.883843] mmc1: no vmmc regulator found
>>> [ 1.895867] mmc_start_host:mmc1: ENTER
>>> [ 1.907217] mmc_detect_change: mmc1: work = df94b950 ENTER
>>> [ 1.926684] NOT SDIO/SD/MMC
>>> [ 1.935001] mmc_schedule_delayed_work: work = df94b950 ENTER
>>> [ 1.951835] mmc_rescan:mmc0 FAIL
>>> [ 1.961423] mmc0: mmc_rescan: trying to init card at 300000 Hz
>>> [ 1.978795] mmc_schedule_delayed_work: x = 1 EXIT
>>> [ 1.992779] mmc_detect_change: mmc1: EXIT
>>> [ 2.004701] mmc1: SDHCI controller on MMC [sdhci-pxa.1] using DMA
>>> [ 2.023003] mmc2: no vmmc regulator found
>>> [ 2.034943] mmc_start_host:mmc2: ENTER
>>> [ 2.049483] NOT SDIO/SD/MMC
>>> [ 2.057800] mmc_detect_change: mmc2: work = df94b550 ENTER
>>> [ 2.074191] mmc_rescan:mmc0 FAIL
>>> [ 2.083788] mmc0: mmc_rescan: trying to init card at 200000 Hz
>>> [ 2.101161] mmc_schedule_delayed_work: work = df94b550 ENTER
>>> [ 2.117985] mmc_schedule_delayed_work: x = 1 EXIT
>>> [ 2.132036] mmc_detect_change: mmc2: EXIT
>>> [ 2.144059] mmc2: SDHCI controller on MMC [sdhci-pxa.2] using DMA
>>> [ 2.162384] TCP cubic registered
>>> [ 2.175181] NOT SDIO/SD/MMC
>>> [ 2.183513] NET: Registered protocol family 17
>>> [ 2.196730] mmc_rescan:mmc0 FAIL
>>> [ 2.206388] Registering the dns_resolver key type
>>> [ 2.220386] mmc0: mmc_rescan: trying to init card at 100000 Hz
>>> [ 2.237749] VFP support v0.3: implementor 56 architecture 2 part 20 variant 9 rev 5
>>> [ 2.262155] max8925-rtc max8925-rtc: setting system clock to 2007-02-05 23:22:44 UTC (1170717764)
>>> [ 2.288769] Waiting 5sec before mounting root device...
>>> [ 2.307747] NOT SDIO/SD/MMC
>>> [ 2.316086] mmc_rescan:mmc0 FAIL
>>> [ 2.325673] mmc_rescan:mmc1: work = df94b950 ENTER
>>> [ 2.339951] mmc1: mmc_rescan: trying to init card at 400000 Hz
>>> [ 2.399680] NOT SDIO/SD/MMC
>>> [ 2.408001] mmc_rescan:mmc1 FAIL
>>> [ 2.417581] mmc1: mmc_rescan: trying to init card at 300000 Hz
>>> [ 2.480967] NOT SDIO/SD/MMC
>>> [ 2.489284] mmc_rescan:mmc1 FAIL
>>> [ 2.498867] mmc1: mmc_rescan: trying to init card at 200000 Hz
>>> [ 2.563544] NOT SDIO/SD/MMC
>>> [ 2.563551] mmc_rescan:mmc1 FAIL
>>> [ 2.581423] mmc1: mmc_rescan: trying to init card at 100000 Hz
>>> [ 2.651318] NOT SDIO/SD/MMC
>>> [ 2.659632] mmc_rescan:mmc1 FAIL
>>> [ 2.669232] mmc_rescan:mmc2: work = df94b550 ENTER
>>> [ 2.683465] mmc2: mmc_rescan: trying to init card at 400000 Hz
>>> [ 2.777696] mmc2: new high speed MMC card at address 0001
>>> [ 2.794089] mmcblk0: mmc2:0001 SEM08G 7.39 GiB
>>> [ 2.813042] mmcblk0: p1
>>> [ 7.325888] EXT3-fs: barriers not enabled
>>> [ 7.398090] EXT3-fs (mmcblk0p1):
>>> [ 7.407518] kjournald starting. Commit interval 5 seconds
>>> [ 7.424361] warning: mounting unchecked fs, running e2fsck is recommended
>>> [ 7.445876] EXT3-fs (mmcblk0p1): using internal journal
>>> [ 7.455234] EXT3-fs (mmcblk0p1): recovery complete
>>> [ 7.476877] EXT3-fs (mmcblk0p1): mounted filesystem with ordered data mode
>>> [ 7.497416] VFS: Mounted root (ext3 filesystem) on device 179:1.
>>> [ 7.515291] Freeing init memory: 92K
>>> init started: BusyBox v1.01 (2009.03.10-14:28+0000) multi-call binary
>>> mount: Mounting none on /tmp failed: No such file or directory
>>> telnetd: daemon: No such file or directory
>>> cannot open /dev/null
>>> [ 7.925389] udevd (57): /proc/57/oom_adj is deprecated, please use /proc/57/oom_score_adj instead.
>>> cannot open /dev/null
>>> error initializing udevd socket
>>> modprobe: could not parse modules.dep
>>>
>>> cannot open /dev/null
>>> error initializing udevd socket
>>> cannot open /dev/null
>>> error initializing udevd socket
>>> cannot open /dev/null
>>> error initializing udevd socket
>>> [ 9.061143] MTD: Attempt to mount non-MTD device "/dev/mmcblk0p1"
>>> device '/dev/ttyS2' does not exist.
>>> [ 9.597226] MTD: Attempt to mount non-MTD device "/dev/mmcblk0p1"
>>>
>>>
>>> BusyBox v1.01 (2009.03.10-14:28+0000) Built-in shell (ash)
>>> Enter 'help' for a list of built-in commands.
>>>
>>>
>>> Processing /etc/profile... ln: /var/dpf/target/: File exists
>>> [ 10.209832] MTD: Attempt to mount non-MTD device "/dev/mmcblk0p1"
>>> error: cannot open /dev/i2c-user
>>> error: cannot open /dev/i2c-user
>>> insmod: /lib/modules/2.6.37-rc5-next-20101208-00003-g4ee4acc-dirty: No such file or directory
>>> Using /lib/modules/galcore.ko
>>> [ 10.498543] galcore: version magic '2.6.28 mod_unload modversions ARMv5 ' should be '2.6.37-rc5-next-20101208-00003-g4ee4acc-dirty preempt mod_unload ARMv6 '
>>> insmod: cannot insert `/lib/modules/galcore.ko': Invalid module format (-1): Exec format error
>>> insmod: /lib/modules/2.6.37-rc5-next-20101208-00003-g4ee4acc-dirty: No such file or directory
>>> insmod: sd8688.ko: no module by that name found
>>> [ 10.927510] MTD: Attempt to mount non-MTD device "/dev/mmcblk0p1"
>>> VERIFY ID: numid=81,iface=MIXER,name=''
>>> amixer: Control default open error: No such device
>>>
>>> VERIFY ID: numid=85,iface=MIXER,name=''
>>> amixer: Control default open error: No such device
>>>
>>> VERIFY ID: numid=1,iface=MIXER,name=''
>>> amixer: Control default open error: No such device
>>>
>>> VERIFY ID: numid=3,iface=MIXER,name=''
>>> amixer: Control default open error: No such device
>>>
>>> VERIFY ID: numid=4,iface=MIXER,name=''
>>> amixer: Control default open error: No such device
>>>
>>> Done
>>>
>>> Philip
>>>
>>
>> --
>> To unsubscribe from this list: send the line "unsubscribe linux-mmc" in
>> the body of a message to majordomo@xxxxxxxxxxxxxxx
>> More majordomo info at http://vger.kernel.org/majordomo-info.html
>>
>

--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at http://vger.kernel.org/majordomo-info.html
Please read the FAQ at http://www.tux.org/lkml/

Next message: Stephen Rothwell: "Re: rcu-walk and dcache scaling tree update and status"
Previous message: Mike Travis: "Early kernel messages are overflowing the static log buffer"
In reply to: Jaehoon Chung: "Re: [BUG]: mmc: single thread workqueue very bad - breaks multiplePERMANENT CARDS"
Messages sorted by: [ date ] [ thread ] [ subject ] [ author ]