Libata error messages

From ata Wiki
Revision as of 08:07, 18 February 2021 by SamSong (Talk | contribs)

(diff) ← Older revision | Latest revision (diff) | Newer revision → (diff)
Jump to: navigation, search



All libata error messages produced by the kernel use a standard format:

ata3.00: exception Emask 0x0 SAct 0x0 SErr 0x0 action 0x2 frozen
ata3.00: cmd ea/00:00:00:00:00/00:00:00:00:00/a0 tag 0
         res 40/00:00:01:4f:c2/00:00:00:00:00/00 Emask 0x4 (timeout)
ata3.00: status: { DRDY }


The prefix


decodes as

ata prefix, indicating this is a libata port or device message
3 port number, counting from one (1)
00 device number, usually zero unless Port Multiplier or PATA master/slave is involved

Exception line

The exception line gives an overview of the EH (Error Handler) state.

exception Emask 0x0 SAct 0x0 SErr 0x0 action 0x2 frozen
Emask Error classification bitmask (AC_ERR_xxx in source code)
SAct SATA SActive register
SErr SATA SError register
action ATA_EH_xxx actions, like revalidate, softreset, hardreset (see include/linux/libata.h)
frozen if present, indicates the port was frozen for EH
t<number> number of retries

Input taskfile

The "cmd" line gives the ATA command (taskfile) sent to the device:

cmd ea/00:00:00:00:00/00:00:00:00:00/a0 tag 0

This lists ATA registers in the following order:

ea Command (FLUSH CACHE EXT EAh, Non-Data)
/ (separator)
00 Feature
00 NSect
00 LBA L
00 LBA M
00 LBA H
/ (separator)
00 HOB Feature
00 HOB NSect
/ (separator)
a0 Device/Head
tag NCQ tag
0 NCQ tag number, or listed as zero if NCQ is not active/applicable.

Output taskfile, error summary

The next line contains a current dump of the ATA device's registers, along with an error summary:

res 40/00:00:01:4f:c2/00:00:00:00:00/00 Emask 0x4 (timeout)

In order:

40 Status
/ (separator)
00 Error
00 NSect
01 LBA L
4f LBA M
c2 LBA H
/ (separator)
00 HOB Error
00 HOB NSect
/ (separator)
00 Device/Head
Emask ATA command's internal error mask (AC_ERR_xxx in source code)
0x4 An English summary of the error, such as
  • timeout
  • HSM violation
  • media error

See below for a full list.

Error classes

These are the possible values for the internal error mask in each error message mentioned above.

AC_ERR_XXX, ATA Completion Errors were defined in include/linux/libata.h.

0x20 host bus error Host<->chip bus error (i.e. PCI, if on PCI bus)
0x10 ATA bus error chip<->device bus error
0x4 timeout Controller failed to respond to an active ATA command. This could be any number of causes. Most often this is due to an unrelated interrupt subsystem bug (try booting with 'pci=nomsi' or 'acpi=off' or 'noapic'), which failed to deliver an interrupt when we were expecting one from the hardware.
0x2 HSM violation Hardware failed to respond in an expected manner. "HSM" stands for Host State Machine, a software-based finite state machine required by ATA that expects certain hardware behaviors, based on the current ATA command and other hardware-state programming details.
0x40 internal error Hardware flagged an impossible condition, most likely due to software misprogramming.
0x8 media error Software detected a media error
0x80 invalid argument Software marked ATA command as invalid, for some reason
0x1 device error Hardware indicates an error with last command. This error is delivered directly from the ATA device. If you see a lot of these, that is often an indication of a hardware problem.
0x100 unknown error Uncategorized error (should never happen)

ATA status expansion

The final line

status: { DRDY }

expands the ATA status register returned in the output taskfile into its component bits:

Busy Device busy (all other bits invalid)
DRDY Device ready. Normally 1, when all is OK.
DRQ Data ready to be sent/received via PIO
DF Device fault
ERR Error (see Error register for more info)

ATA error expansion

If any bits in the Error register are set, the Error register contents will be expanded into its component bits, for example:

error: { ICRC ABRT }
ICRC Interface CRC error during Ultra DMA transfer - often either a bad cable or power problem, though possibly an incorrect Ultra DMA mode setting by the driver
UNC Uncorrectable error - often due to bad sectors on the disk
IDNF Requested address was not found
ABRT Command aborted - either command not supported, unable to complete, or interface CRC (with ICRC)

SATA SError expansion

If any bits in the SATA SError register are set, the SError register contents will be expanded into its component bits, for example:

SError: { PHYRdyChg CommWake }

These bits are set by the SATA host interface in response to error conditions on the SATA link. Unless a drive hotplug or unplug operation occurred, it is generally not normal to see any of these bits set. If they are, it usually points strongly toward a hardware problem (often a bad SATA cable or a bad or inadequate power supply).

RecovData Data integrity error occurred, but the interface recovered
RecovComm Communications between device and host temporarily lost, but regained
UnrecovData Data integrity error occurred, interface did not recover
Persist Persistent communication or data integrity error
Proto SATA protocol violation detected
HostInt Host bus adapter internal error
PHYRdyChg PhyRdy signal changed state
PHYInt PHY internal error
CommWake COMWAKE detected by PHY (PHY woken up)
10B8B 10b to 8b decoding error occurred
Dispar Incorrect disparity detected
BadCRC Link layer CRC error occurred
Handshk R_ERR handshake response received in response to frame transmission
LinkSeq Link state machine error occurred
TrStaTrns Transport layer state transition error occurred
UnrecFIS Unrecognized FIS (frame information structure) received
DevExch Device presence has changed
Personal tools