Skip to contentSkip to conversation

Samsung 840 EVO

DISCUSSION Günter Waller 3 participants 11 messages May 4, 2015 libata

lore
Share & tools
lore

Günter Waller

author
#0 lore raw

Dear Tejun,
the subject Samsung SSD is having performance issues with "old data" which Samsung now addresses with a new firmware. As often is the case, Linux users find it a bit more difficult to apply that fix. However, I also have a Windows system, so it should be no big deal for me.

But there is also a big concern for Linux users which this FW version seems to introduce. The link below points to a warning (backed by others, elsewhere on the web) that fstrim in combination with this FW will wreck your drive.
http://www.overclock.net/t/1507897/samsung-840-evo-read-speed-drops-on-old-written-data-in-the-drive/2640%23post_23827674
(search for fstrim)
I raised this in a discussion thread of the Ubuntu MATE community and was pointed to the SSD blacklist that is part of the module libata-core.c which names you as the maintainer. I wonder if you could take a look at this issue - maybe you want to add the device to the blacklist thereafter if need be.
If you find the time and find this worthwhile to investigate I would appreciate a brief notification of the outcome. For now I am holding back the FW update.

Best regards

Guenter Waller

#1 lore raw
Quoted 1 line
Quoted 1 line
Quoted 1 line
Quoted 1 line
Quoted 1 line

"Günter" == Günter Waller <[email protected]> writes:

Günter> the subject Samsung SSD is having performance issues with "old Günter> data" which Samsung now addresses with a new firmware.

I wonder if they enabled the same botched queued discard that they did in 850 Pro?

If someone has an 840 Evo with the new firmware it would be helpful for us to receive the output of:

# hdparm -I /dev/sdN
# smartctl -l gplog,0x13 /dev/sdN

I think I might have an 840 in the lab. Will head down and update it later this week unless somebody beats me to providing the information above.

Signature · 1 line
Martin K. Petersen	Oracle Linux Engineering
#2 lore raw

Am 04.05.2015 um 18:28 schrieb Martin K. Petersen:

Quoted 5 lines

If someone has an 840 Evo with the new firmware it would be helpful for us to receive the output of:

# hdparm -I /dev/sdN
# smartctl -l gplog,0x13 /dev/sdN

Find the output attached. But I'm not sure the smartctl is helpful, since my machine has been running with discard disabled for about half a day now.

/dev/sda:

ATA device, with non-removable media

	Model Number:       Samsung SSD 840 EVO 250GB               
	Serial Number:      deleted
	Firmware Revision:  EXT0DB6Q
	Transport:          Serial, ATA8-AST, SATA 1.0a, SATA II Extensions, SATA Rev 2.5, SATA Rev 2.6, SATA Rev 3.0
Standards:
	Used: unknown (minor revision code 0x0039) 
	Supported: 9 8 7 6 5 
	Likely used: 9

Configuration:
Logical max current
cylinders 16383 16383
heads 16 16
sectors/track 63 63
--
CHS current addressable sectors: 16514064
LBA user addressable sectors: 268435455
LBA48 user addressable sectors: 488397168
Logical Sector size: 512 bytes
Physical Sector size: 512 bytes
Logical Sector-0 offset: 0 bytes
device size with M = 1024*1024: 238475 MBytes
device size with M = 1000*1000: 250059 MBytes (250 GB)
cache/buffer size = unknown
Nominal Media Rotation Rate: Solid State Device
Capabilities:
LBA, IORDY(can be disabled)
Queue depth: 32
Standby timer values: spec'd by Standard, no device specific minimum
R/W multiple sector transfer: Max = 1 Current = 1
DMA: mdma0 mdma1 mdma2 udma0 udma1 udma2 udma3 udma4 udma5 *udma6
Cycle time: min=120ns recommended=120ns
PIO: pio0 pio1 pio2 pio3 pio4
Cycle time: no flow control=120ns IORDY flow control=120ns
Commands/features:
Enabled Supported:
* SMART feature set
Security Mode feature set
* Power Management feature set
* Write cache
* Look-ahead
* Host Protected Area feature set
* WRITE_BUFFER command
* READ_BUFFER command
* NOP cmd
* DOWNLOAD_MICROCODE
SET_MAX security extension
* 48-bit Address feature set
* Device Configuration Overlay feature set
* Mandatory FLUSH_CACHE
* FLUSH_CACHE_EXT
* SMART error logging
* SMART self-test
* General Purpose Logging feature set
* WRITE_{DMA|MULTIPLE}_FUA_EXT
* 64-bit World wide name
Write-Read-Verify feature set
* WRITE_UNCORRECTABLE_EXT command
* {READ,WRITE}_DMA_EXT_GPL commands
* Segmented DOWNLOAD_MICROCODE
* Gen1 signaling speed (1.5Gb/s)
* Gen2 signaling speed (3.0Gb/s)
* Gen3 signaling speed (6.0Gb/s)
* Native Command Queueing (NCQ)
* Phy event counters
* unknown 76[15]
* DMA Setup Auto-Activate optimization
Device-initiated interface power management
* Asynchronous notification (eg. media change)
* Software settings preservation
* SMART Command Transport (SCT) feature set
* SCT LBA Segment Access (AC2)
* SCT Error Recovery Control (AC3)
* SCT Features Control (AC4)
* SCT Data Tables (AC5)
* reserved 69[4]
* DOWNLOAD MICROCODE DMA command
* SET MAX SETPASSWORD/UNLOCK DMA commands
* WRITE BUFFER DMA command
* READ BUFFER DMA command
* Data Set Management TRIM supported (limit 8 blocks)
Security:
Master password revision code = 65534
supported
not enabled
not locked
frozen
not expired: security count
supported: enhanced erase
2min for SECURITY ERASE UNIT. 8min for ENHANCED SECURITY ERASE UNIT.
Logical Unit WWN Device Identifier: 50025388a05456c8

	NAA		: 5
	IEEE OUI	: 002538
	Unique ID	: 8a05456c8

Checksum: correct

smartctl 6.3 2014-07-26 r3976 [x86_64-linux-3.19.6] (local build) Copyright (C) 2002-14, Bruce Allen, Christian Franke, www.smartmontools.org

General Purpose Log 0x13 [SATA NCQ Send and Receive log], Page 0-0 (of 1)

0000000: 01 00 00 00 01 00 00 00 00 00 00 00 00 00 00 00 |................|
0000010: 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 |................|
0000020: 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 |................|
0000030: 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 |................|
0000040: 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 |................|
0000050: 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 |................|
0000060: 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 |................|
0000070: 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 |................|
0000080: 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 |................|
0000090: 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 |................|
00000a0: 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 |................|
00000b0: 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 |................|
00000c0: 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 |................|
00000d0: 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 |................|
00000e0: 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 |................|
00000f0: 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 |................|
0000100: 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 |................|
0000110: 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 |................|
0000120: 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 |................|
0000130: 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 |................|
0000140: 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 |................|
0000150: 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 |................|
0000160: 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 |................|
0000170: 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 |................|
0000180: 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 |................|
0000190: 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 |................|
00001a0: 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 |................|
00001b0: 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 |................|
00001c0: 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 |................|
00001d0: 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 |................|
00001e0: 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 |................|
00001f0: 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 |................|
#3 lore raw
Quoted 3 lines
Quoted 3 lines
Quoted 1 line
Quoted 1 line
Quoted 1 line

"Sven" == Sven Köhler <[email protected]> writes:

# hdparm -I /dev/sdN # smartctl -l gplog,0x13 /dev/sdN

Sven> I'm not sure the smartctl is helpful, since my machine has been Sven> running with discard disabled for about half a day now.

It is helpful. It shows us that Samsung has indeed enabled queued TRIM functionality in this firmware update.

Signature · 1 line
Martin K. Petersen	Oracle Linux Engineering
#4 lore raw

Am 04.05.2015 um 19:21 schrieb Martin K. Petersen:

Quoted 9 lines
Quoted 3 lines
Quoted 3 lines
Quoted 1 line
Quoted 1 line
Quoted 1 line

"Sven" == Sven Köhler <[email protected]> writes:

# hdparm -I /dev/sdN # smartctl -l gplog,0x13 /dev/sdN

Sven> I'm not sure the smartctl is helpful, since my machine has been Sven> running with discard disabled for about half a day now.

It is helpful. It shows us that Samsung has indeed enabled queued TRIM functionality in this firmware update.

Are you saying that previous firmwares didn't claim that the drive supports queued TRIM. But the new firmware does - even though the feature appears to be broken?

Regards,
Sven

#5 lore raw
Quoted 4 lines
Quoted 4 lines
Quoted 1 line
Quoted 1 line
Quoted 1 line

"Sven" == Sven Köhler <[email protected]> writes:

It is helpful. It shows us that Samsung has indeed enabled queued TRIM functionality in this firmware update.

Sven> Are you saying that previous firmwares didn't claim that the drive Sven> supports queued TRIM. But the new firmware does - even though the Sven> feature appears to be broken?

Correct. The old firmware releases for this drive did not advertise support for queued TRIM whereas the new one does. We saw exactly the same issue with a firmware update for 850 Pro.

Signature · 1 line
Martin K. Petersen	Oracle Linux Engineering
author
#6 lore raw
In reply to Martin K. Petersen

Hi Martin,
here you are - the second command, however, did not seem to produce anything useful.
But beware: this is the old firmware, EXT0BB6Q, i.e. the one with the performance issues (not seen on my machine yet), not the one that causes the problems with fstrim - named EXT0DB6Q.
As I mentioned I shy away from upgrading at this point, not knowing whether my Linux system runs fstrim somewhere behind my back. I am not that deeply involved in the inner workings of the kernel as you guys. This is my bread and butter machine, I'd rather not expose it to risk if I can help it.
Cheers
Günter

Am 04.05.2015 um 17:28 schrieb Martin K. Petersen:

Quoted 11 lines
Quoted 1 line
Quoted 1 line
Quoted 1 line
Quoted 1 line
Quoted 1 line

"Günter" == Günter Waller <[email protected]> writes:

Günter> the subject Samsung SSD is having performance issues with "old Günter> data" which Samsung now addresses with a new firmware.

I wonder if they enabled the same botched queued discard that they did in 850 Pro?

If someone has an 840 Evo with the new firmware it would be helpful for us to receive the output of:

Quoted 7 lines
# hdparm -I /dev/sdN
# smartctl -l gplog,0x13 /dev/sdN

I think I might have an 840 in the lab. Will head down and update it later this week unless somebody beats me to providing the information above.

hdparm -I /dev/sda
/dev/sda:

ATA device, with non-removable media

	Model Number:       Samsung SSD 840 EVO 250GB               
	Serial Number:      S1DBNSBF767383T     
	Firmware Revision:  EXT0BB6Q
	Transport:          Serial, ATA8-AST, SATA 1.0a, SATA II Extensions, SATA Rev 2.5, SATA Rev 2.6, SATA Rev 3.0
Standards:
	Used: unknown (minor revision code 0x0039) 
	Supported: 9 8 7 6 5 
	Likely used: 9

Configuration:
Logical max current
cylinders 16383 16383
heads 16 16
sectors/track 63 63
--
CHS current addressable sectors: 16514064
LBA user addressable sectors: 268435455
LBA48 user addressable sectors: 488397168
Logical Sector size: 512 bytes
Physical Sector size: 512 bytes
Logical Sector-0 offset: 0 bytes
device size with M = 1024*1024: 238475 MBytes
device size with M = 1000*1000: 250059 MBytes (250 GB)
cache/buffer size = unknown
Nominal Media Rotation Rate: Solid State Device
Capabilities:
LBA, IORDY(can be disabled)
Queue depth: 32
Standby timer values: spec'd by Standard, no device specific minimum
R/W multiple sector transfer: Max = 1 Current = 1
DMA: mdma0 mdma1 mdma2 udma0 udma1 udma2 udma3 udma4 udma5 *udma6
Cycle time: min=120ns recommended=120ns
PIO: pio0 pio1 pio2 pio3 pio4
Cycle time: no flow control=120ns IORDY flow control=120ns
Commands/features:
Enabled Supported:
* SMART feature set
Security Mode feature set
* Power Management feature set
* Write cache
* Look-ahead
* Host Protected Area feature set
* WRITE_BUFFER command
* READ_BUFFER command
* NOP cmd
* DOWNLOAD_MICROCODE
SET_MAX security extension
* 48-bit Address feature set
* Device Configuration Overlay feature set
* Mandatory FLUSH_CACHE
* FLUSH_CACHE_EXT
* SMART error logging
* SMART self-test
* General Purpose Logging feature set
* WRITE_{DMA|MULTIPLE}_FUA_EXT
* 64-bit World wide name
Write-Read-Verify feature set
* WRITE_UNCORRECTABLE_EXT command
* {READ,WRITE}_DMA_EXT_GPL commands
* Segmented DOWNLOAD_MICROCODE
* Gen1 signaling speed (1.5Gb/s)
* Gen2 signaling speed (3.0Gb/s)
* Gen3 signaling speed (6.0Gb/s)
* Native Command Queueing (NCQ)
* Phy event counters
* READ_LOG_DMA_EXT equivalent to READ_LOG_EXT
* DMA Setup Auto-Activate optimization
Device-initiated interface power management
* Asynchronous notification (eg. media change)
* Software settings preservation
* SMART Command Transport (SCT) feature set
* SCT Write Same (AC2)
* SCT Error Recovery Control (AC3)
* SCT Features Control (AC4)
* SCT Data Tables (AC5)
* reserved 69[4]
* DOWNLOAD MICROCODE DMA command
* SET MAX SETPASSWORD/UNLOCK DMA commands
* WRITE BUFFER DMA command
* READ BUFFER DMA command
* Data Set Management TRIM supported (limit 8 blocks)
Security:
Master password revision code = 65534
supported
not enabled
not locked
frozen
not expired: security count
supported: enhanced erase
2min for SECURITY ERASE UNIT. 8min for ENHANCED SECURITY ERASE UNIT.
Logical Unit WWN Device Identifier: 50025388a05d2afa

	NAA		: 5
	IEEE OUI	: 002538
	Unique ID	: 8a05d2afa

Checksum: correct

*************************************************************
smartctl -l gplog,0x13 /dev/sda
smartctl 6.2 2013-07-26 r3841 [x86_64-linux-3.13.0-51-generic] (local build) Copyright (C) 2002-13, Bruce Allen, Christian Franke, www.smartmontools.org

General Purpose Log 0x13 does not exist (override with '-T permissive' option)

#7 lore raw
Quoted 1 line
Quoted 1 line
Quoted 1 line
Quoted 1 line
Quoted 1 line

"Günter" == Günter Waller <[email protected]> writes:

Günter> the second command, however, did not seem to produce anything Günter> useful. But beware: this is the old firmware, EXT0BB6Q,

Yes, this confirms that the old firmware did not support queued TRIM. Whereas the new one does.

I'll try to reach out to my contacts at Samsung again so we can get this resolved.

Until then, I have sent out a patch that blacklists queued TRIM for all Samsung 800-series drives. It appears to be generic to their implementation and not tied to a particular drive model.

Signature · 1 line
Martin K. Petersen	Oracle Linux Engineering
author
#8 lore raw

Martin,
do you mind if I post the essence of your last two mails to my thread in the Ubuntu MATE forum? So the maintainers of that system can see what is going on.

Am 04.05.2015 um 18:24 schrieb Martin K. Petersen:

Quoted 15 lines
Quoted 1 line
Quoted 1 line
Quoted 1 line
Quoted 1 line
Quoted 1 line

"Günter" == Günter Waller <[email protected]> writes:

Günter> the second command, however, did not seem to produce anything Günter> useful. But beware: this is the old firmware, EXT0BB6Q,

Yes, this confirms that the old firmware did not support queued TRIM. Whereas the new one does.

I'll try to reach out to my contacts at Samsung again so we can get this resolved.

Until then, I have sent out a patch that blacklists queued TRIM for all Samsung 800-series drives. It appears to be generic to their implementation and not tied to a particular drive model.

#9 lore raw
Quoted 1 line
Quoted 1 line
Quoted 1 line
Quoted 1 line
Quoted 1 line

"Günter" == Günter Waller <[email protected]> writes:

Günter> Martin, do you mind if I post the essence of your last two mails Günter> to my thread in the Ubuntu MATE forum?

Go right ahead.

Signature · 1 line
Martin K. Petersen	Oracle Linux Engineering
#10 lore raw
In reply to Günter Waller

Am 04.05.2015 um 15:38 schrieb Günter Waller:

Quoted 6 lines

But there is also a big concern for Linux users which this FW version seems to introduce. The link below points to a warning (backed by others, elsewhere on the web) that fstrim in combination with this FW will wreck your drive.
http://www.overclock.net/t/1507897/samsung-840-evo-read-speed-drops-on-old-written-data-in-the-drive/2640%23post_23827674
(search for fstrim)

Quoted 8 lines

I raised this in a discussion thread of the Ubuntu MATE community and was pointed to the SSD blacklist that is part of the module libata-core.c which names you as the maintainer. I wonder if you could take a look at this issue - maybe you want to add the device to the blacklist thereafter if need be.
If you find the time and find this worthwhile to investigate I would appreciate a brief notification of the outcome. For now I am holding back the FW update.

Here is a discussion of the issue and a patch:
https://bugs.launchpad.net/ubuntu/+source/fstrim/+bug/1449005

Apparently, TRIM in combination with NCQ is broken with the new firmware. Having the discard mount option enabled didn't wreck my drive, but it kept my machine from booting. I had to disable that option and the patch given at the URL above seems to disable NCQ specifically for TRIM only. So the performance hit might be minimal.

The patch should really be applied asap and it should be backported to all stable branched as far as possible.

Regards,
Sven