Self tests pass under Windows 10 1809 but get stuck at 90% under Linux

[Artem S. Tashkinov]

I've got a strange issue with my SSD disk: all self tests pass under Windows 10 1809 but get stuck at 90% under Linux no matter how much I wait for their completion. I barely have any disk activity, so there's no reason for the tests to get stuck.

Under both OSes I use the same smartmontools version - 7.0 release.

In the attached log 90% "aborted by host" are Linux results. 00% completed

[Artem S. Tashkinov]

... 00% "completed without error" are Windows results.

[Artem S. Tashkinov]

This has been happening for quite some time under Linux kernels 4.18, 4.19, 5.0 and 5.1. Haven't tested Linux 5.2 because it hasn't been made available in Fedora 30 yet. I'm not sure any previous kernels worked at all.

This looks like a weird bug in the Linux kernel SATA layer but I'm an absolute newbie in this area, so it's just a wild guess.

[Artem S. Tashkinov]

Also, I'm curious why I see so many errors.

[Christian Franke]

... I barely have any disk activity, so there's no reason for the tests to get stuck.

The missing disk activity may actually be the reason for aborted self-tests. See the FAQ for explanation and possible workaround.

[Christian Franke]

Also, I'm curious why I see so many errors.

This is unrelated to the above.

ID# ATTRIBUTE_NAME          FLAGS    VALUE WORST THRESH FAIL RAW_VALUE
...
  5 Reallocated_Sector_Ct   PO--CK   100   100   010    -    0
  9 Power_On_Hours          -O--CK   092   092   001    -    4987
...
187 Uncorrectable_Error_Cnt -O--CK   100   100   000    -    5
...
196 Reallocated_Event_Count -O--CK   253   253   001    -    0
198 Offline_Uncorrectable   ----CK   100   100   001    -    0

No reallocated sectors, no sectors pending for reallocation, ...

SMART Extended Comprehensive Error Log Version: 1 (1 sectors)
...
Error 5 [0] occurred at disk power-on lifetime: 3665 hours (152 days + 17 hours)
  When the command that caused the error occurred, the device was active or idle.

  After command completion occurred, registers were:
  ER -- ST COUNT  LBA_48  LH LM LL DV DC
  -- -- -- == -- == == == -- -- -- -- --
  40 -- 51 00 40 00 00 00 00 00 14 40 00  Error: UNC at LBA = 0x00000014 = 20

  Commands leading to the command that caused the error were:
  CR FEATR COUNT  LBA_48  LH LM LL DV DC  Powered_Up_Time  Command/Feature_Name
  -- == -- == -- == == == -- -- -- -- --  ---------------  --------------------
  60 00 01 00 40 00 00 00 00 00 14 40 08     00:02:31.000  READ FPDMA QUEUED
...
Error 4 [3] occurred at disk power-on lifetime: 3525 hours (146 days + 21 hours)
...
  40 -- 51 00 20 00 00 00 00 00 10 40 00  Error: UNC at LBA = 0x00000010 = 16
...
Error 3 [2] occurred at disk power-on lifetime: 3483 hours (145 days + 3 hours)
...
  40 -- 51 00 40 00 00 00 00 00 14 40 00  Error: UNC at LBA = 0x00000014 = 20
...
Error 2 [1] occurred at disk power-on lifetime: 3348 hours (139 days + 12 hours)
...
  40 -- 51 00 40 00 00 00 00 00 14 40 00  Error: UNC at LBA = 0x00000014 = 20
...

... very few transient read errors on LBA 16 and 20 occurred more than 1300 power on hours ago, ...

SMART Extended Self-test Log Version: 1 (1 sectors)
Num  Test_Description    Status                  Remaining  LifeTime(hours)  LBA_of_first_error
# 1  Extended offline    Completed without error       00%      4987         -

... but a recent extended test succeeded. Conclusion: No sign of trouble. The possible weak sectors have been fixed by new write commands.

PS: This is a bug tracker, not a support forum. For future support questions, please use the ​smartmontools-support mailing list instead.

[Artem S. Tashkinov]

The missing disk activity may actually be the reason for aborted self-tests. See the FAQ for explanation and possible workaround.

This is definitely not the case. It's a system disk, so there's some activity.

What I meant with "next to no activity" is that I'm not copying huge files between partitions or anything like that.

[Christian Franke]

Check kernel log for any disk related messages which occur during self-test.

Check these counters before and after self-test:

SATA Phy Event Counters (GP Log 0x11)
ID      Size     Value  Description
...
0x0009  2            2  Transition from drive PhyRdy to drive PhyNRdy
0x000a  2            2  Device-to-host register FISes sent due to a COMRESET

If there is any increase, there must have been an event which resets the SATA link or the device. This typically aborts any self-test.

[Artem S. Tashkinov]

The issue persists even with kernel 5.16.19.

No idea what's going on, nothing is logged in dmesg, all the tests invariably fail with "Interrupted (host reset)" even if I run smartctl -a hours after I run smartctl -t long/short.

It's as if tests immediately fail when launched under Linux.

[Artem S. Tashkinov]

Before and after running smartct -t short /dev/sda which should have completed in 2 minutes. The second smartctl -x was run 5 minutes after attempting to run a short self-test. The test was aborted at 90%.

.txt

@@ -32 +32 @@
 Offline data collection status:  (0x82) Offline data collection activity
                                         was completed without error.
                                         Auto Offline Data Collection: Enabled.
-Self-test execution status:      (   0) The previous self-test routine completed
-                                        without error or no self-test has ever 
-                                        been run.
+Self-test execution status:      ( 249) Self-test routine in progress...
+                                        90% of test remaining.
 Total time to complete Offline 
 data collection:                ( 1800) seconds.
 Offline data collection
@@ -73 +72 @@
 184 End-to-End_Error        PO--CK   100   100   097    -    0
 187 Uncorrectable_Error_Cnt -O--CK   100   100   000    -    5
 188 Command_Timeout         -O--CK   100   100   000    -    0
-190 Airflow_Temperature_Cel -O---K   062   045   040    -    38
+190 Airflow_Temperature_Cel -O---K   059   045   040    -    41
 196 Reallocated_Event_Count -O--CK   253   253   001    -    0
 198 Offline_Uncorrectable   ----CK   100   100   001    -    0
 199 CRC_Error_Count         -O---K   100   100   001    -    0
@@ -213 +212 @@
 
 Warning! SMART Selective Self-Test Log Structure error: invalid SMART checksum.
 SMART Selective self-test log data structure revision number 1
- SPAN  MIN_LBA  MAX_LBA  CURRENT_TEST_STATUS
-    1        0       10  Not_testing
-    2        0        0  Not_testing
-    3        0        0  Not_testing
-    4        0        0  Not_testing
-    5        0        0  Not_testing
-  255        0    65535  Read_scanning was completed without error
+ SPAN    MIN_LBA    MAX_LBA  CURRENT_TEST_STATUS
+    1          0         10  Not_testing
+    2          0          0  Not_testing
+    3          0          0  Not_testing
+    4          0          0  Not_testing
+    5          0          0  Not_testing
+  255  247475376  247540911  Read_scanning was completed without error
 Selective self-test flags (0x0):
   After scanning selected spans, do NOT read-scan remainder of disk.
 If Selective self-test is pending on power-up, resume after 0 minute delay.
@@ -229 +228 @@
 Device Statistics (GP Log 0x04)
 Page  Offset Size        Value Flags Description
 0x01  =====  =               =  ===  == General Statistics (rev 2) ==
-0x01  0x018  6      9158167139  ---  Logical Sectors Written
-0x01  0x028  6      9286355418  ---  Logical Sectors Read
-0x01  0x1f8  7               0  NDC+ Unknown
+0x01  0x018  6      9158258099  ---  Logical Sectors Written
+0x01  0x028  6      9286355426  ---  Logical Sectors Read
 0x05  =====  =               =  ===  == Temperature Statistics (rev 1) ==
-0x05  0x008  1              38  ---  Current Temperature
+0x05  0x008  1              41  ---  Current Temperature
 0x05  0x020  1              50  ---  Highest Temperature
 0x05  0x028  1              26  ---  Lowest Temperature
                                 |||_ C monitored condition met
@@ -252 +250 @@
 0x0006  2            0  R_ERR response for device-to-host non-data FIS
 0x0007  2            0  R_ERR response for host-to-device non-data FIS
 0x0008  2            0  Device-to-host non-data FIS retries
-0x0009  2            0  Transition from drive PhyRdy to drive PhyNRdy
+0x0009  2            1  Transition from drive PhyRdy to drive PhyNRdy
 0x000a  2            2  Device-to-host register FISes sent due to a COMRESET
 0x000b  2            0  CRC errors within host-to-device FIS
 0x000d  2            0  Non-CRC errors within host-to-device FIS

[Artem S. Tashkinov]

There's absolutely nothing in dmesg.

[asm]

I had the same problem yesterday in Debian 12 stable (bookworm) running kernel

$ uname -a
Linux XXXXXXXXXXXXXXXXXXX 6.1.0-27-amd64 #1 SMP PREEMPT_DYNAMIC Debian 6.1.115-1 (2024-11-01) x86_64 GNU/Linux

smartctl output is:

...<snip>
Model Family:     Samsung based SSDs
Device Model:     Samsung SSD 840 EVO 1TB
...<snip>
SMART Self-test log structure revision number 1
Num  Test_Description    Status                  Remaining  LifeTime(hours)  LBA_of_first_error
...<snip>
# 6  Extended offline    Aborted by host               00%        13         -
...<snip>

However, I started smartctl via the GUI gsmartctl. The problem could be related to gsmartctl. In the next few days I will run the same extended tests again with the command line program smartcl directly. I hope that the test will then run without stopping at 90%. If so, then it is due to gsmartctl and not smartctl, if not, then smartctl or somewhere in the kernel is the problem.

With short self-tests, the time is usually too short for the bug to be reliably triggered.