hdparm on nvme ssd super slow (100 MB/s)

Message

Gh0str1d3r · Post by **Gh0str1d3r** » Fri Aug 31, 2018 8:19 pm

The SystemRescueCd live USB gives the same slow speed of the SSD:

/dev/nvme0n1:
 Timing O_DIRECT cached reads:   428 MB in  2.00 seconds = 213.52 MB/sec
 Timing O_DIRECT disk reads: 516 MB in  3.00 seconds = 171.85 MB/sec

If Windows wouldn't be faster I'd say the device is broken. But like this is seems to be a Linux issue.

For comparison, the result of lspci -nkk in the live USB environment:

Code: Select all

00:00.0 0600: 8086:1904 (rev 08)
	Subsystem: 1028:0704
	Kernel driver in use: skl_uncore
00:02.0 0300: 8086:1916 (rev 07)
	Subsystem: 1028:0704
	Kernel driver in use: i915
	Kernel modules: i915
00:04.0 1180: 8086:1903 (rev 08)
	Subsystem: 1028:0704
	Kernel driver in use: proc_thermal
	Kernel modules: processor_thermal_device
00:14.0 0c03: 8086:9d2f (rev 21)
	Subsystem: 1028:0704
	Kernel driver in use: xhci_hcd
00:14.2 1180: 8086:9d31 (rev 21)
	Subsystem: 1028:0704
	Kernel driver in use: intel_pch_thermal
	Kernel modules: intel_pch_thermal
00:15.0 1180: 8086:9d60 (rev 21)
	Subsystem: 1028:0704
	Kernel driver in use: intel-lpss
	Kernel modules: intel_lpss_pci
00:15.1 1180: 8086:9d61 (rev 21)
	Subsystem: 1028:0704
	Kernel driver in use: intel-lpss
	Kernel modules: intel_lpss_pci
00:16.0 0780: 8086:9d3a (rev 21)
	Subsystem: 1028:0704
	Kernel driver in use: mei_me
	Kernel modules: mei_me
00:17.0 0106: 8086:9d03 (rev 21)
	Subsystem: 1028:0704
	Kernel driver in use: ahci
00:1c.0 0604: 8086:9d10 (rev f1)
	Kernel driver in use: pcieport
	Kernel modules: shpchp
00:1c.4 0604: 8086:9d14 (rev f1)
	Kernel driver in use: pcieport
	Kernel modules: shpchp
00:1c.5 0604: 8086:9d15 (rev f1)
	Kernel driver in use: pcieport
	Kernel modules: shpchp
00:1d.0 0604: 8086:9d18 (rev f1)
	Kernel driver in use: pcieport
	Kernel modules: shpchp
00:1f.0 0601: 8086:9d48 (rev 21)
	Subsystem: 1028:0704
00:1f.2 0580: 8086:9d21 (rev 21)
	Subsystem: 1028:0704
	Kernel driver in use: intel_pmc_core
00:1f.3 0403: 8086:9d70 (rev 21)
	Subsystem: 1028:0704
00:1f.4 0c05: 8086:9d23 (rev 21)
	Subsystem: 1028:0704
	Kernel driver in use: i801_smbus
	Kernel modules: i2c_i801
3a:00.0 0280: 8086:095a (rev 59)
	Subsystem: 8086:5010
	Kernel driver in use: iwlwifi
	Kernel modules: iwlwifi
3b:00.0 ff00: 10ec:525a (rev 01)
	Subsystem: 1028:0704
	Kernel driver in use: rtsx_pci
	Kernel modules: rtsx_pci
3c:00.0 0108: 144d:a802 (rev 01)
	Subsystem: 144d:a801
	Kernel driver in use: nvme
	Kernel modules: nvme

The only issue I could find is an extremely high number of Error Log Entries:

Code: Select all

# smartctl -a /dev/nvme0     
smartctl 6.6 2017-11-05 r4594 [x86_64-linux-4.18.3-gentoo] (local build)
Copyright (C) 2002-17, Bruce Allen, Christian Franke, www.smartmontools.org

=== START OF INFORMATION SECTION ===
Model Number:                       PM951 NVMe SAMSUNG 1024GB
Serial Number:                      S2FZNXAG900687
Firmware Version:                   BXV76D0Q
PCI Vendor/Subsystem ID:            0x144d
IEEE OUI Identifier:                0x002538
Controller ID:                      1
Number of Namespaces:               1
Namespace 1 Size/Capacity:          1,024,209,543,168 [1.02 TB]
Namespace 1 Utilization:            935,313,330,176 [935 GB]
Namespace 1 Formatted LBA Size:     512
Namespace 1 IEEE EUI-64:            002538 45591b02af
Local Time is:                      Fri Aug 31 16:18:29 2018 EDT
Firmware Updates (0x06):            3 Slots
Optional Admin Commands (0x0017):   Security Format Frmw_DL Self_Test
Optional NVM Commands (0x001f):     Comp Wr_Unc DS_Mngmt Wr_Zero Sav/Sel_Feat
Maximum Data Transfer Size:         32 Pages

Supported Power States
St Op     Max   Active     Idle   RL RT WL WT  Ent_Lat  Ex_Lat
 0 +     6.00W       -        -    0  0  0  0        5       5
 1 +     4.20W       -        -    1  1  1  1       30      30
 2 +     3.10W       -        -    2  2  2  2      100     100
 3 -   0.0700W       -        -    3  3  3  3      500    5000
 4 -   0.0050W       -        -    4  4  4  4     2000   22000

Supported LBA Sizes (NSID 0x1)
Id Fmt  Data  Metadt  Rel_Perf
 0 +     512       0         0

=== START OF SMART DATA SECTION ===
SMART overall-health self-assessment test result: PASSED

SMART/Health Information (NVMe Log 0x02, NSID 0xffffffff)
Critical Warning:                   0x00
Temperature:                        42 Celsius
Available Spare:                    100%
Available Spare Threshold:          50%
Percentage Used:                    1%
Data Units Read:                    48,210,029 [24.6 TB]
Data Units Written:                 21,591,699 [11.0 TB]
Host Read Commands:                 669,617,398
Host Write Commands:                534,740,380
Controller Busy Time:               3,647
Power Cycles:                       2,969
Power On Hours:                     7,495
Unsafe Shutdowns:                   162
Media and Data Integrity Errors:    0
Error Information Log Entries:      10,090

Error Information (NVMe Log 0x01, max 64 entries)
Num   ErrCount  SQId   CmdId  Status  PELoc          LBA  NSID    VS
  0      10090     0  0x001b  0x4004  0x000            0     0     -
  1      10089     0  0x0013  0x4212  0x000            0   255     -
  2      10088     0  0x0051  0x4004  0x000            0     0     -
  3      10087     0  0x0050  0x4004  0x000            0     0     -
  4      10086     0  0x004f  0x4004  0x000            0     0     -
  5      10085     0  0x004e  0x4004  0x000            0     0     -
  6      10084     0  0x0015  0x4004  0x000            0     0     -
  7      10083     0  0x0014  0x4004  0x000            0     0     -
  8      10082     0  0x006a  0x4004  0x000            0     0     -
  9      10081     0  0x00e4  0x4004  0x000            0     0     -
 10      10080     0  0x0060  0x4004  0x000            0     0     -
 11      10079     0  0x00a6  0x4004  0x000            0     0     -
 12      10078     0  0x00d6  0x4004  0x000            0     0     -
 13      10077     0  0x00cd  0x4004  0x000            0     0     -
 14      10076     0  0x00f3  0x4004  0x000            0     0     -
 15      10075     0  0x0017  0x4004  0x000            0     0     -
... (48 entries not shown)

Code: Select all

# nvme smart-log /dev/nvme0n1
Smart Log for NVME device:nvme0n1 namespace-id:ffffffff
critical_warning                    : 0
temperature                         : 41 C
available_spare                     : 100%
available_spare_threshold           : 50%
percentage_used                     : 1%
data_units_read                     : 48,210,030
data_units_written                  : 21,591,751
host_read_commands                  : 669,617,435
host_write_commands                 : 534,740,950
controller_busy_time                : 3,647
power_cycles                        : 2,969
power_on_hours                      : 7,495
unsafe_shutdowns                    : 162
media_errors                        : 0
num_err_log_entries                 : 10,090
Warning Temperature Time            : 0
Critical Composite Temperature Time : 0
Thermal Management T1 Trans Count   : 0
Thermal Management T2 Trans Count   : 0
Thermal Management T1 Total Time    : 0
Thermal Management T2 Total Time    : 0

Does anyone know me what this means?

russK · Post by **russK** » Fri Aug 31, 2018 9:36 pm

Gh0str1d3r,

It looks like the device believes it is healthy and has had no errors, which is good to know.

sysrescue does not have some drivers, like Anon-E-moose and NeddySeagoon were helping you solve for your gentoo kernel. If that's still also the case in your gentoo boot, there is still more to solve here, since the device is fast as expected in Windows. Unfortunately sysrescue did not load them by default, maybe you could modprobe them and see if the speed changes.

Have you noticed this wiki page: https://wiki.gentoo.org/wiki/Dell_XPS_13_9350
It does not talk about nvme specifically but it may show the special sauce you need.
Also Arch wiki has a similar page for that laptop, although there may be many variants.

HTH

Blind_Sniper · Post by **Blind_Sniper** » Fri Aug 31, 2018 9:54 pm

I would try to use other utility, just to make sure that hdparm is correct, e.g. dd command:

write speed:

Code: Select all

dd if=/dev/zero of=/device/mount/point/output.img bs=8k count=256k

Then clean cache:

Code: Select all

sync; echo 1 > /proc/sys/vm/drop_caches
sync; echo 2 > /proc/sys/vm/drop_caches
sync; echo 3 > /proc/sys/vm/drop_caches

and finally read speed:

Code: Select all

dd if=/device/mount/point/output.img of=/dev/null bs=8k count=256k

will it be the same values like you got from hdparm?

As for hdd, it gives comparable results (dd vs hdparm)

added: make sure you have enough free space at the drive ( 2Gb )

Gh0str1d3r · Post by **Gh0str1d3r** » Sat Sep 01, 2018 8:37 am

I now have all drivers, thanks for the tip with the live USB! But the ssd is still as slow as it was before. The dd command runs a little faster, thanks for the suggestion!

Code: Select all

root@dell-laptop # hdparm -Tt --direct /dev/nvme0n1

/dev/nvme0n1:
 Timing O_DIRECT cached reads:   544 MB in  2.01 seconds = 271.23 MB/sec
 Timing O_DIRECT disk reads: 662 MB in  3.00 seconds = 220.49 MB/sec

Code: Select all

root@dell-laptop # dd if=/dev/zero of=/root/aa.img bs=8k count=256k

262144+0 records in
262144+0 records out
2147483648 bytes (2.1 GB, 2.0 GiB) copied, 17.8105 s, 121 MB/s

Code: Select all

root@dell-laptop # sync; echo 1 > /proc/sys/vm/drop_caches; sync; echo 2 > /proc/sys/vm/drop_caches ;sync; echo 3 > /proc/sys/vm/drop_caches

Code: Select all

root@dell-laptop # dd if=/root/aa.img of=/dev/null bs=8k count=256k

262144+0 records in
262144+0 records out
2147483648 bytes (2.1 GB, 2.0 GiB) copied, 2.6398 s, 814 MB/s

But first this is still not the ~2.5-3.5 GB/s that is specified and second everyone is using 'hdparm -tT --direct' to test the speed of this drive. Why would mine not work?

P.Kosunen · Post by **P.Kosunen** » Mon Sep 03, 2018 11:27 am

Code: Select all

nvme error-log /dev/nvme0n1

What kind of errors are there?

Gh0str1d3r · Post by **Gh0str1d3r** » Mon Sep 03, 2018 6:07 pm

P.Kosunen wrote:
Code: Select all
nvme error-log /dev/nvme0n1
What kind of errors are there?

https://pastebin.com/0KAXPM5v

I don't know how to interpret this.

P.Kosunen · Post by **P.Kosunen** » Tue Sep 04, 2018 1:34 pm

Can you test if these errors are coming on Windows, Gentoo or both sides?

russK · Post by **russK** » Sat Sep 08, 2018 4:21 am

Sorry I did not notice the number of error log entries a 4 or 5 days ago.

I don't know what it means, but my device only has an error count of 5. I will try to keep watching this thread.

Regards