Errors in memory regression test

Hi All,
I was doing memory regression test for our customized TK1 board.
When DVFS regression was executed and lasted about 8 hours, system will print below messages.

dvfs logging end runtime: 8:00:00.091988

DVFS done for 8:00:00! Launching DFS task.

SD2          ON       1300.0mV
LDO0         ON       1000.0mV
SD1          ON       1250.0mV
SD1          ON       1230.0mV
SD1          ON       1210.0mV
SD1          ON       1190.0mV
SD1          ON       1170.0mV
SD1          ON       1150.0mV
SD1          ON       1130.0mV
SD1          ON       1110.0mV
SD1          ON       1090.0mV
SD1          ON       1070.0mV
SD1          ON       1050.0mV
SD1          ON       1030.0mV
SD1          ON       1010.0mV
SD1          ON        990.0mV
SD1          ON        980.0mV
SD1          ON       1250.0mV
SD1          ON       1230.0mV
SD1          ON       1210.0mV
SD1          ON       1190.0mV
SD1          ON       1170.0mV
....
....
SD1          ON       1050.0mV
SD1          ON       1030.0mV
SD1          ON       1010.0mV
SD1          ON        990.0mV
SD1          ON        970.0mV
SD1          ON        950.0mV
SD1          ON        930.0mV
[32823.268048] mmc0: Data CRC error, intmask: 200000
[32823.272764] sdhci: =========== REGISTER DUMP (mmc0)===========
[32823.278600] sdhci: Sys addr: 0x00000000 | Version:  0x00000303
[32823.284434] sdhci: Blk size: 0x00007200 | Blk cnt:  0x0000001e
[32823.290267] sdhci: Argument: 0x00c85188 | Trn mode: 0x00000023
[32823.296105] sdhci: Present:  0x01fb00f0 | Host ctl: 0x00000031
[32823.301941] sdhci: Power:    0x0000000b | Blk gap:  0x00000000
[32823.307777] sdhci: Wake-up:  0x00000000 | Clock:    0x00000007
[32823.313611] sdhci: Timeout:  0x0000000e | Int stat: 0x00000000
[32823.319444] sdhci: Int enab: 0x02ff000b | Sig enab: 0x02fc000b
[32823.325271] sdhci: AC12 err: 0x00000000 | Slot int: 0x00000000
[32823.331101] sdhci: Caps:     0x376fd080 | Caps_1:   0x10002f77
[32823.336928] sdhci: Cmd:      0x0000193a | Max curr: 0x00000000
[32823.342756] sdhci: Host ctl2: 0x0000308b
[32823.346677] sdhci: ADMA Err: 0x00000000 | ADMA Ptr: 0xae840010
[32823.352501] sdhci: ===========================================
[32823.359148] mmcblk0: error -84 transferring data, sector 13128072, nr 32, cmd response 0x900, card status 0xc00
[32823.369425] end_request: I/O error, dev mmcblk0, sector 13128072
[32823.375564] end_request: I/O error, dev mmcblk0, sector 13128080
[32823.381743] end_request: I/O error, dev mmcblk0, sector 13128088
[32823.387853] end_request: I/O error, dev mmcblk0, sector 13128096
[32823.394070] Aborting journal on device mmcblk0p1-8.
[32823.400436] journal commit I/O error
[32823.401122] EXT4-fs error (device mmcblk0p1): __ext4_journal_start_sb:62: Detected aborted journal
[32823.401132] EXT4-fs (mmcblk0p1): Remounting filesystem read-only
Traceback (most recent call last):
  File "/home/memtest/temp_logger.py", line 16, in <module>
OSError: [Errno 5] Input/output error
SD1          ON        910.0mV
Traceback (most recent call last):
  File "/usr/lib/python2.7/logging/__init__.py", line 875, in emit
    self.flush()
  File "/usr/lib/python2.7/logging/__init__.py", line 835, in flush
    self.stream.flush()
IOError: [Errno 30] Read-only file system
Logged from file l4t_dfs_switch.py, line 225
SD1          ON        890.0mV
Traceback (most recent call last):
  File "/usr/lib/python2.7/logging/__init__.py", line 875, in emit
    self.flush()
  File "/usr/lib/python2.7/logging/__init__.py", line 835, in flush
    self.stream.flush()
IOError: [Errno 30] Read-only file system
Logged from file l4t_dfs_switch.py, line 225
SD1          ON        870.0mV
Traceback (most recent call last):
  File "/usr/lib/python2.7/logging/__init__.py", line 875, in emit
    self.flush()
  File "/usr/lib/python2.7/logging/__init__.py", line 835, in flush
    self.stream.flush()
IOError: [Errno 30] Read-only file system
Logged from file l4t_dfs_switch.py, line 225
SD1          ON        850.0mV
Traceback (most recent call last):
  File "/home/memtest/run_nvmmt.py", line 12, in <module>
OSError: [Errno 30] Read-only file system
Traceback (most recent call last):
  File "/usr/lib/python2.7/logging/__init__.py", line 875, in emit
    self.flush()
  File "/usr/lib/python2.7/logging/__init__.py", line 835, in flush
    self.stream.flush()
IOError: [Errno 30] Read-only file system
Logged from file l4t_dfs_switch.py, line 225
Traceback (most recent call last):
  File "/usr/lib/python2.7/logging/__init__.py", line 875, in emit
    self.flush()
  File "/usr/lib/python2.7/logging/__init__.py", line 835, in flush
    self.stream.flush()
IOError: [Errno 30] Read-only file system
...
...

It seemed emmc became read-only because CRC error.
I have tried this on Jetson TK1 board and got the same issue.

Any idea?
Thanks.

Hi, what dvfs table are you using on TK1 board? If the dvfs table is generated from your custom board, it should not be used on TK1 board.

Hi Trumany,
I didn’t use the customized dvfs table on Jetson TK1 board. I just flashed official L4t r21.6 image on Jestson TK1 board and execute the DVFS memory regression test. Have you tried regression test on Jetson TK1?
The command I test is

python l4t_TK1_V01.py -b pm375 -p ams -t dvfs_2nvmmt_island -f 924,792,600,528,396,300,204

Please refer to Note in page 51: Please check if eMMC is working in DDR50 mode; Vcore may violate eMMC HS200 Vmin limitation when static regression is performed.

Hi Trumany,
I ran below commands before doing DVFS memory regression test.

$ sudo bash
$ echo 4 > /sys/kernel/debug/mmc0/speed

The test worked properly with no errors.
Thanks for your help.