ExpressCard compact flash card very slow

When using a Lexar ExpressCard CompactFlash (CF) reader in an ExpressCard slot, the speed of this high speed device is very slow (relatively speaking, of course).It performs at ~28MB/s instead of ~80MB/s it should be with a fast CompactFlash card.

This is the product: http://www.lexar.com/products/lexar-professional-expresscard-compactflash-cf-reader?category=213

The CF card can run at UDMA/133, but is set to run at UDMA/33.

$dmesg

[ 1.687119] scsi0 : pata_jmicron
[ 1.692480] scsi1 : pata_jmicron
[ 1.692523] ata1: PATA max UDMA/100 cmd 0x4020 ctl 0x4014 bmdma 0x4000 irq 19
[ 1.692525] ata2: PATA max UDMA/100 cmd 0x4018 ctl 0x4010 bmdma 0x4008 irq 19
[ 1.870088] usb 1-6: new high speed USB device using ehci_hcd and address 3
[ 1.890676] ata1.00: CFA: TRANSCEND, 20091215, max UDMA/133
[ 1.890678] ata1.00: 125059072 sectors, multi 0: LBA
[ 1.890681] ata1.00: limited to UDMA/33 due to 40-wire cable
[ 1.930831] ata1.00: configured for UDMA/33
[ 1.930984] scsi 0:0:0:0: Direct-Access ATA TRANSCEND 2009 PQ: 0 ANSI: 5
[ 1.931132] sd 0:0:0:0: [sda] 125059072 512-byte logical blocks: (64.0 GB/59.6 GiB)
[ 1.931144] sd 0:0:0:0: Attached scsi generic sg0 type 0
[ 1.931181] sd 0:0:0:0: [sda] Write Protect is off
[ 1.931184] sd 0:0:0:0: [sda] Mode Sense: 00 3a 00 00
[ 1.931371] sd 0:0:0:0: [sda] Write cache: disabled, read cache: enabled, doesn't support DPO or FUA
[ 1.932525] sda: sda1
[ 1.932904] sd 0:0:0:0: [sda] Attached SCSI disk

$sudo lspci -vvv -xxx

05:00.0 IDE interface: JMicron Technology Corp. JMB368 IDE controller (prog-if 85 [Master SecO PriO])
Subsystem: JMicron Technology Corp. JMB368 IDE controller
Control: I/O+ Mem+ BusMaster+ SpecCycle- MemWINV- VGASnoop- ParErr-
Stepping- SERR+ FastB2B- DisINTx-
Status: Cap+ 66MHz- UDF- FastB2B- ParErr- DEVSEL=fast >TAbort-
<TAbort- <MAbort- >SERR- <PERR- INTx-

    Latency: 0, Cache Line Size: 64 bytes
    Interrupt: pin A routed to IRQ 19
    Region 0: I/O ports at 4020 [size=8]
    Region 1: I/O ports at 4014 [size=4]
    Region 2: I/O ports at 4018 [size=8]
    Region 3: I/O ports at 4010 [size=4]
    Region 4: I/O ports at 4000 [size=16]
    [virtual] Expansion ROM at f2900000 [disabled] [size=64K]
    Capabilities: [68] Power Management version 2
        Flags: PMEClk- DSI- D1- D2- AuxCurrent=0mA 

PME(D0-,D1-,D2-,D3hot-,D3cold-)
Status: D0 NoSoftRst- PME-Enable- DSel=0 DScale=0 PME-
Capabilities: [50] Express (v1) Legacy Endpoint, MSI 01
DevCap: MaxPayload 128 bytes, PhantFunc 0, Latency L0s
<64ns, L1 <1us
ExtTag- AttnBtn- AttnInd- PwrInd- RBE- FLReset-
DevCtl: Report errors: Correctable- Non-Fatal- Fatal-
Unsupported-
RlxdOrd- ExtTag- PhantFunc- AuxPwr- NoSnoop- MaxPayload 128 bytes, MaxReadReq 512 bytes
DevSta: CorrErr- UncorrErr- FatalErr- UnsuppReq+ AuxPwr-
TransPend-
LnkCap: Port #1, Speed 2.5GT/s, Width x1, ASPM L0s, Latency
L0 <1us, L1 <16us
ClockPM- Surprise- LLActRep- BwNot-
LnkCtl: ASPM Disabled; RCB 64 bytes Disabled- Retrain- CommClk+
ExtSynch- ClockPM- AutWidDis- BWInt- AutBWInt-
LnkSta: Speed 2.5GT/s, Width x1, TrErr- Train- SlotClk+
DLActive- BWMgmt- ABWMgmt-
Capabilities: [70] MSI: Enable- Count=1/1 Maskable- 64bit-
Address: 00000000 Data: 0000
Kernel driver in use: pata_jmicron
Kernel modules: pata_jmicron
00: 7b 19 68 23 07 01 10 00 00 85 01 01 10 00 00 00 10: 21 40 00 00 15 40 00 00 19 40 00 00 11 40 00 00 20: 01 40 00 00 00 00 00 00 00 00 00 00 7b 19 68 23 30: 00 00 00 00 68 00 00 00 00 00 00 00 0b 01 00 00 40: bd 00 c0 81 a8 08 80 e0 00 00 00 00 00 00 00 00 50: 10 70 11 02 00 00 00 00 00 20 08 00 11 44 02 01 60: 40 00 11 10 00 00 00 00 01 50 02 00 00 00 00 00 70: 05 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 80: 00 00 00 00 a0 01 00 00 0f 00 00 00 00 00 00 00 90: 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00
a0: 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00
b0: 31 00 00 00 00 00 00 08 00 00 00 00 00 00 00 00
c0: c3 8b 38 70 03 45 02 00 00 00 00 00 00 00 00 00
d0: 18 00 00 80 00 00 00 10 02 40 eb 00 00 00 00 01
e0: 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00
f0: 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00

$ sudo hdparm -I /dev/sda

/dev/sda:

CompactFlash ATA device

    Model Number:       TRANSCEND
    Serial Number:      20100521    000008F4
    Firmware Revision:  20091215

Standards:
Likely used: 6
Configuration:
Logical max current
cylinders 16383 65535
heads 15 15
sectors/track 63 63
--
CHS current addressable sectors: 61930575
LBA user addressable sectors: 125059072
Logical/Physical Sector size: 512 bytes
device size with M = 1024*1024: 61064 MBytes
device size with M = 1000*1000: 64030 MBytes (64 GB)
cache/buffer size = 1 KBytes (type=DualPort)
Capabilities:
LBA, IORDY(can be disabled)
bytes avail on r/w long: 4
Standby timer values: spec'd by Vendor
R/W multiple sector transfer: Max = 1 Current = 0
Advanced power management level: disabled
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
WRITE_BUFFER command
READ_BUFFER command
NOP cmd
CFA feature set
Advanced Power Management feature set Mandatory FLUSH_CACHE
* CFA advanced modes: pio5 *pio6 mdma3 mdma4
* CFA Power Level 1 (max 500mA)
Security:
Master password revision code = 65534
supported
not enabled
not locked
not frozen
not expired: security count
not supported: enhanced erase
6min for SECURITY ERASE UNIT.
HW reset results:
CBLID- above Vih
Device num = 0
Integrity word not set (found 0x0000, expected 0x5da5)

$ sudo hdparm --Istdout /dev/sda

/dev/sda:
044a 3fff 0000 000f 0000 0240 003f 0774
4000 0000 3230 3130 3035 3231 2020 2020
3030 3030 3038 4634 0002 0002 0004 3230
3039 3132 3135 5452 414e 5343 454e 4420
2020 2020 2020 2020 2020 2020 2020 2020
2020 2020 2020 2020 2020 2020 2020 8001
0000 0f00 0000 0200 0000 0007 ffff 000f
003f fc4f 03b0 0100 4000 0774 0000 0007
0003 0078 0078 0078 0078 0000 0000 0000
0000 0000 0000 0000 0000 0000 0000 0000
0000 0000 702b 500c 4003 0001 0000 0003
047f 0003 0000 0000 fffe 604f 0000 0000
0000 0000 0000 0000 0000 0000 0000 0000
0000 0000 0000 0000 0000 0000 0000 0000
0000 0000 0000 0000 0000 0000 0000 0000
0000 0000 0000 0000 0000 0000 0000 0000
0001 0000 0000 0000 0000 0000 0000 0000
0000 0000 0000 0000 0000 0000 0000 0000
0000 0000 0000 0000 0000 0000 0000 0000
0000 0000 0000 0000 0000 0000 0000 0000
81f4 0000 0000 0092 0000 0000 0000 0000
0000 0000 0000 0000 0000 0000 0000 0000
0000 0000 0000 0000 0000 0000 0000 0000
0000 0000 0000 0000 0000 0000 0000 0000
0000 0000 0000 0000 0000 0000 0000 0000
0000 0000 0000 0000 0000 0000 0000 0000
0000 0000 0000 0000 0000 0000 0000 0000
0000 0000 0000 0000 0000 0000 0000 0000
0000 0000 0000 0000 0000 0000 0000 0000
0000 0000 0000 0000 0000 0000 0000 0000
0000 0000 0000 0000 0000 0000 0000 0000
0000 0000 0000 0000 0000 0000 0000 0000

Upstream kernel, on Ubuntu 11.04 alpha:

$ cat /proc/version
Linux version 2.6.38-020638rc6-generic (root@zinc) (gcc version 4.2.3 (Ubuntu 4.2.3-2ubuntu7)) #201102220910 SMP Tue Feb 22 09:12:52 UTC 2011

The CF adapter probably doesn't support the 80-wire cable detect protocol properly so the kernel thinks there's only a 40-wire cable connected to it. You can try libata.force=80c on the kernel command line and see what that does.

It seemed to work. I got read speeds of ~55MB/s, which is ok because I was writing what was being read to an SSD.

dmesg:

[    1.651748] pata_jmicron 0000:05:00.0: PCI INT A -> GSI 19 (level, low) -> IRQ 19
[    1.651783] pata_jmicron 0000:05:00.0: setting latency timer to 64
[    1.652212] scsi0 : pata_jmicron
[    1.652813] scsi1 : pata_jmicron
[    1.652855] ata1: PATA max UDMA/100 cmd 0x4020 ctl 0x4014 bmdma 0x4000 irq 19
[    1.652857] ata2: PATA max UDMA/100 cmd 0x4018 ctl 0x4010 bmdma 0x4008 irq 19
[    1.850716] ata1: FORCE: cable set to 80c
[    1.850723] ata1.00: CFA: TRANSCEND, 20091215, max UDMA/133
[    1.850726] ata1.00: 125059072 sectors, multi 0: LBA
[    1.870111] usb 1-6: new high speed USB device using ehci_hcd and address 3
[    1.890684] ata1.00: configured for UDMA/100
[    1.890845] scsi 0:0:0:0: Direct-Access     ATA      TRANSCEND        2009 PQ: 0 ANSI: 5
[    1.891008] sd 0:0:0:0: Attached scsi generic sg0 type 0
[    1.891020] sd 0:0:0:0: [sda] 125059072 512-byte logical blocks: (64.0 GB/59.6 GiB)
[    1.891070] sd 0:0:0:0: [sda] Write Protect is off
[    1.891073] sd 0:0:0:0: [sda] Mode Sense: 00 3a 00 00
[    1.891099] sd 0:0:0:0: [sda] Write cache: disabled, read cache: enabled, doesn't support DPO or FUA
[    1.892022]  sda: sda1
[    1.892395] sd 0:0:0:0: [sda] Attached SCSI disk

What next? Can the CF reader be whitelisted by libata?

2011/2/23 Damon Lynch <[email protected]>:

I don't think thats a safe idea, as not all CF adaptors are created equally, so I could imagine that you could easily break alot of existing machines doing so...

The CF reader portion itself likely isn't identifiable (it's probably mostly just passive circuitry). The PATA adapter portion might be identifiable, unless it just has a totally generic JMicron device ID.. What does lspci -vvnn show for it?

05:00.0 IDE interface [0101]: JMicron Technology Corp. JMB368 IDE controller [197b:2368] (prog-if 85 [Master SecO PriO])
Subsystem: JMicron Technology Corp. JMB368 IDE controller [197b:2368]
Control: I/O+ Mem+ BusMaster+ SpecCycle- MemWINV- VGASnoop- ParErr-
Stepping- SERR+ FastB2B- DisINTx-
Status: Cap+ 66MHz- UDF- FastB2B- ParErr- DEVSEL=fast >TAbort-
<TAbort- <MAbort- >SERR- <PERR- INTx-

    Latency: 0, Cache Line Size: 64 bytes
    Interrupt: pin A routed to IRQ 19
    Region 0: I/O ports at 4020 [size=8]
    Region 1: I/O ports at 4014 [size=4]
    Region 2: I/O ports at 4018 [size=8]
    Region 3: I/O ports at 4010 [size=4]
    Region 4: I/O ports at 4000 [size=16]
    [virtual] Expansion ROM at f2900000 [disabled] [size=64K]
    Capabilities: [68] Power Management version 2
        Flags: PMEClk- DSI- D1- D2- AuxCurrent=0mA 

PME(D0-,D1-,D2-,D3hot-,D3cold-)
Status: D0 NoSoftRst- PME-Enable- DSel=0 DScale=0 PME-
Capabilities: [50] Express (v1) Legacy Endpoint, MSI 01
DevCap: MaxPayload 128 bytes, PhantFunc 0, Latency L0s
<64ns, L1 <1us
ExtTag- AttnBtn- AttnInd- PwrInd- RBE- FLReset-
DevCtl: Report errors: Correctable- Non-Fatal- Fatal-
Unsupported-
RlxdOrd- ExtTag- PhantFunc- AuxPwr- NoSnoop- MaxPayload 128 bytes, MaxReadReq 512 bytes
DevSta: CorrErr- UncorrErr- FatalErr- UnsuppReq+ AuxPwr-
TransPend-
LnkCap: Port #1, Speed 2.5GT/s, Width x1, ASPM L0s, Latency
L0 <1us, L1 <16us
ClockPM- Surprise- LLActRep- BwNot-
LnkCtl: ASPM Disabled; RCB 64 bytes Disabled- Retrain- CommClk+
ExtSynch- ClockPM- AutWidDis- BWInt- AutBWInt-
LnkSta: Speed 2.5GT/s, Width x1, TrErr- Train- SlotClk+
DLActive- BWMgmt- ABWMgmt-
Capabilities: [70] MSI: Enable- Count=1/1 Maskable- 64bit-
Address: 00000000 Data: 0000
Kernel driver in use: pata_jmicron
Kernel modules: pata_jmicron

Yeah, it looks like it just has a generic subsystem/subvendor ID so we can't tell this thing apart from any other JMicron JMB368 PATA controller. Not sure there is anything we can do here to detect this case automatically.

Though, I suppose one other thing we could do is just ignore cable detection results for CompactFlash devices - it seems like many CF adapters get this wrong. If the user really connected a fast CF card with a 40-wire cable, the error handling would probably do the right thing and lower the transfer mode if errors occurred.