First of all: A big "Thank You!" especially to Sarah Sharp and Hans de Goede. Finding their contributions finally led my way. #################################################################################################################### HARDWARE involved: ################## - Xeon E3-1276 v3 - Asus P9D WS - - 00:14.0 USB controller [0c03] : - - Intel Corporation 8 Series/C220 Series Chipset Family USB xHCI [8086:8c31] (rev 05) - LC-Power "LC-35U3-Becrux" with - - ASM1153E soldered onto its adapter board <-- <-- - - - 130926 a1 00 00 original firmware version - - - 140509_A1_82_80.bin newest firmware update I could gather from - - - http://plugable.com/drivers/hddockfirmware - Seagate Barracuda 7200.11 1.5TB (ST315003 41AS) #################################################################################################################### Kernel versions involved: ######################### - linux-4.1.12-gentoo <-- failure, as described below - linux-4.1.13-gentoo <-- failure - linux-4.2.6-gentoo <-- failure - linux-4.3.0-gentoo <-- failure - linux-4.3.0-gentoo with "usb-storage.quirks=174c:55aa:u" <-- working [Workaround] #################################################################################################################### SHORT STORY: ############ ASM continues shuffling new HW under the old hat of 174c:55aa. Connecting the external enclosure, lsusb displays an "ASM1051 SATA 3Gb/s bridge" - you need a microscope to finally detect the chip as an ASM1153E instead. Usage of the device @ USB 2 succeeds. "Normal" usage of the device @ USB 3 succeeds, but under stress the connection fails *instantaneusly* . Adding "usb-storage.quirks=174c:55aa:u" to the kernel start parameters makes the device fall back from UAS to usb-storage, but deliver stable full throughput the disc is capable of. Reproducible: Always #################################################################################################################### LONG STORY: ############ [ 5155.065761] usb 2-6: new SuperSpeed USB device number 2 using xhci_hcd [ 5155.077463] usb 2-6: New USB device found, idVendor=174c, idProduct=55aa [ 5155.077467] usb 2-6: New USB device strings: Mfr=2, Product=3, SerialNumber=1 [ 5155.077469] usb 2-6: Product: mkn_LC-53U3-Becrux [ 5155.077470] usb 2-6: Manufacturer: mkn_LC-POWER [ 5155.077472] usb 2-6: SerialNumber: 201401000056 [ 5155.079306] scsi host7: uas [ 5155.163597] scsi 7:0:0:0: Direct-Access ST315003 41AS 0 PQ: 0 ANSI: 6 [ 5155.163966] sd 7:0:0:0: Attached scsi generic sg6 type 0 [ 5155.164117] sd 7:0:0:0: [sdd] 2930277168 512-byte logical blocks: (1.50 TB/1.36 TiB) [ 5155.164433] sd 7:0:0:0: [sdd] Write Protect is off [ 5155.164435] sd 7:0:0:0: [sdd] Mode Sense: 43 00 00 00 [ 5155.164553] sd 7:0:0:0: [sdd] Write cache: enabled, read cache: enabled, doesn't support DPO or FUA [ 5155.229663] sdd: sdd1 [ 5155.230384] sd 7:0:0:0: [sdd] Attached SCSI disk ### Until here, everything is o.k. ### Afterwards, "normal" / "slow" access to the disk works o.k. ### But when stressing the disk (even without mounting a fs, e.g. with "dd"), ### a READ ERROR occurs right away: [ 7505.098336] sd 7:0:0:0: [sdd] tag#0 data cmplt err -71 uas-tag 1 inflight: CMD [ 7505.098341] sd 7:0:0:0: [sdd] tag#0 CDB: Read(10) 28 00 0e df 25 d0 00 01 00 00 [ 7505.101696] xhci_hcd 0000:00:14.0: ERROR Transfer event for disabled endpoint or incorrect stream ring <-- !!! [ 7505.101708] xhci_hcd 0000:00:14.0: @00000000ffffc6d0 00000000 00000000 04000000 01038000 [ 7535.253307] sd 7:0:0:0: [sdd] tag#1 uas_eh_abort_handler 0 uas-tag 2 inflight: CMD IN [ 7535.253311] sd 7:0:0:0: [sdd] tag#1 CDB: Read(10) 28 00 0e df 26 d0 00 01 00 00 [ 7535.253343] sd 7:0:0:0: [sdd] tag#0 uas_eh_abort_handler 0 uas-tag 1 inflight: CMD [ 7535.253344] sd 7:0:0:0: [sdd] tag#0 CDB: Read(10) 28 00 0e df 25 d0 00 01 00 00 [ 7535.253363] scsi host7: uas_eh_bus_reset_handler start [ 7535.355461] usb 2-6: reset SuperSpeed USB device number 2 using xhci_hcd [ 7535.367384] scsi host7: uas_eh_bus_reset_handler success # From now on, the connection to the disk is lost: [ 7542.276188] sd 7:0:0:0: [sdd] tag#0 FAILED Result: hostbyte=DID_OK driverbyte=DRIVER_SENSE [ 7542.276193] sd 7:0:0:0: [sdd] tag#0 Sense Key : Not Ready [current] [ 7542.276196] sd 7:0:0:0: [sdd] tag#0 Add. Sense: Medium not present <-- !!! [ 7542.276198] sd 7:0:0:0: [sdd] tag#0 CDB: Read(10) 28 00 0e df 26 d0 00 01 00 00 [ 7542.276200] blk_update_request: I/O error, dev sdd, sector 249505488 ### < ... iterated ... > ### Power-OFF / Power-ON the USB Disk: [12629.885763] usb 2-6: new SuperSpeed USB device number 3 using xhci_hcd [12629.897417] usb 2-6: New USB device found, idVendor=174c, idProduct=55aa [12629.897422] usb 2-6: New USB device strings: Mfr=2, Product=3, SerialNumber=1 [12629.897424] usb 2-6: Product: mkn_LC-53U3-Becrux [12629.897426] usb 2-6: Manufacturer: mkn_LC-POWER [12629.897428] usb 2-6: SerialNumber: 201401000056 [12629.898921] scsi host8: uas [12629.976381] scsi 8:0:0:0: Direct-Access ST315003 41AS 0 PQ: 0 ANSI: 6 [12629.976686] sd 8:0:0:0: Attached scsi generic sg6 type 0 [12629.976788] sd 8:0:0:0: [sde] 2930277168 512-byte logical blocks: (1.50 TB/1.36 TiB) [12629.977105] sd 8:0:0:0: [sde] Write Protect is off [12629.977108] sd 8:0:0:0: [sde] Mode Sense: 43 00 00 00 [12629.977222] sd 8:0:0:0: [sde] Write cache: enabled, read cache: enabled, doesn't support DPO or FUA [12630.035392] sde: sde1 [12630.036222] sd 8:0:0:0: [sde] Attached SCSI disk [12634.107292] usb 2-6: USB disconnect, device number 3 [12634.107829] sd 8:0:0:0: [sde] Synchronizing SCSI cache [12634.199253] sd 8:0:0:0: [sde] Synchronize Cache(10) failed: Result: hostbyte=DID_ERROR driverbyte=DRIVER_OK [12634.405237] usb 2-6: new SuperSpeed USB device number 4 using xhci_hcd [12634.416877] usb 2-6: New USB device found, idVendor=174c, idProduct=55aa [12634.416882] usb 2-6: New USB device strings: Mfr=2, Product=3, SerialNumber=1 [12634.416885] usb 2-6: Product: mkn_LC-53U3-Becrux [12634.416888] usb 2-6: Manufacturer: mkn_LC-POWER [12634.416890] usb 2-6: SerialNumber: 201401000056 [12634.418494] scsi host9: uas [12634.418899] scsi 9:0:0:0: Direct-Access ST315003 41AS 0 PQ: 0 ANSI: 6 [12634.419429] sd 9:0:0:0: Attached scsi generic sg6 type 0 [12634.419514] sd 9:0:0:0: [sde] 2930277168 512-byte logical blocks: (1.50 TB/1.36 TiB) [12634.419840] sd 9:0:0:0: [sde] Write Protect is off [12634.419847] sd 9:0:0:0: [sde] Mode Sense: 43 00 00 00 [12634.419999] sd 9:0:0:0: [sde] Write cache: enabled, read cache: enabled, doesn't support DPO or FUA [12634.447456] sde: sde1 [12634.448233] sd 9:0:0:0: [sde] Attached SCSI disk ### NOTABENE: Now it's a 'new' disk _sde_ , not _sdd_ any more ### Stressing disc access with "dd if=/dev/sde of=/dev/zero" results in same failure right away: [13492.949113] sd 9:0:0:0: [sde] tag#0 data cmplt err -71 uas-tag 1 inflight: CMD [13492.949117] sd 9:0:0:0: [sde] tag#0 CDB: Read(10) 28 00 00 08 c6 e0 00 01 00 00 [13492.952418] xhci_hcd 0000:00:14.0: ERROR Transfer event for disabled endpoint or incorrect stream ring <-- !!! [13492.952430] xhci_hcd 0000:00:14.0: @00000000ffffc320 00000000 00000000 04000000 03038001 [13523.257665] sd 9:0:0:0: [sde] tag#1 uas_eh_abort_handler 0 uas-tag 2 inflight: CMD IN [13523.257669] sd 9:0:0:0: [sde] tag#1 CDB: Read(10) 28 00 00 08 c7 e0 00 01 00 00 [13523.257700] sd 9:0:0:0: [sde] tag#0 uas_eh_abort_handler 0 uas-tag 1 inflight: CMD [13523.257702] sd 9:0:0:0: [sde] tag#0 CDB: Read(10) 28 00 00 08 c6 e0 00 01 00 00 [13523.257721] scsi host9: uas_eh_bus_reset_handler start [13523.359827] usb 2-6: reset SuperSpeed USB device number 4 using xhci_hcd [13523.372038] scsi host9: uas_eh_bus_reset_handler success [13530.309911] sd 9:0:0:0: [sde] tag#0 FAILED Result: hostbyte=DID_OK driverbyte=DRIVER_SENSE [13530.309917] sd 9:0:0:0: [sde] tag#0 Sense Key : Not Ready [current] [13530.309919] sd 9:0:0:0: [sde] tag#0 Add. Sense: Medium not present <-- !!! [13530.309921] sd 9:0:0:0: [sde] tag#0 CDB: Read(10) 28 00 00 08 c7 e0 00 01 00 00 [13530.309923] blk_update_request: I/O error, dev sde, sector 575456 ### < ... iterated ... > #-#################################################################################################################### #-### #-### found --> http://thread.gmane.org/gmane.linux.usb.general/119107 #-### and --> linux-.../drivers/usb/host/xhci-ring.c #################################################################################################################### WORKAROUND : ############ ### Modified /boot/grub/grub.conf : ### - - - linux /vmlinuz-4.3.0-gentoo ... usb-storage.quirks=174c:55aa:u <-- !!! [ 79.892893] usb 2-6: new SuperSpeed USB device number 2 using xhci_hcd [ 79.904640] usb 2-6: New USB device found, idVendor=174c, idProduct=55aa [ 79.904648] usb 2-6: New USB device strings: Mfr=2, Product=3, SerialNumber=1 [ 79.904653] usb 2-6: Product: mkn_LC-53U3-Becrux [ 79.904657] usb 2-6: Manufacturer: mkn_LC-POWER [ 79.904660] usb 2-6: SerialNumber: 201401000056 [ 79.905574] usb 2-6: UAS is blacklisted for this device, using usb-storage instead <-- !!! [ 79.906001] usb 2-6: UAS is blacklisted for this device, using usb-storage instead <-- !!! [ 79.906010] usb-storage 2-6:1.0: USB Mass Storage device detected [ 79.906147] usb-storage 2-6:1.0: Quirks match for vid 174c pid 55aa: c00000 <-- !!! [ 79.906215] scsi host7: usb-storage 2-6:1.0 [ 80.965177] scsi 7:0:0:0: Direct-Access ST315003 41AS 0 PQ: 0 ANSI: 6 [ 80.965294] sd 7:0:0:0: Attached scsi generic sg6 type 0 [ 80.965475] sd 7:0:0:0: [sdd] 2930277168 512-byte logical blocks: (1.50 TB/1.36 TiB) [ 80.965657] sd 7:0:0:0: [sdd] Write Protect is off [ 80.965658] sd 7:0:0:0: [sdd] Mode Sense: 43 00 00 00 [ 80.965858] sd 7:0:0:0: [sdd] Write cache: enabled, read cache: enabled, doesn't support DPO or FUA [ 80.990135] sdd: sdd1 [ 80.991087] sd 7:0:0:0: [sdd] Attached SCSI disk [ 180.683067] usb 2-6: reset SuperSpeed USB device number 2 using xhci_hcd ### Stressing disc access with "dd if=/dev/sde of=/dev/zero" : ### ### SUCCESS : on-going transfer rates of ~ 125 MB/s right from the beginning ### which is the maximum a Seagate Barracuda 7200.11 1.5TB (ST31500341AS) can supply. ### ### NO failure any more : ### the whole disk (all of the 1.5 TB) was "read" (dd) successfully in one go. ### But now in dmesg : [ 180.683067] usb 2-6: reset SuperSpeed USB device number 2 using xhci_hcd [ 199.004752] usb 2-6: reset SuperSpeed USB device number 2 using xhci_hcd [ 206.422149] usb 2-6: reset SuperSpeed USB device number 2 using xhci_hcd [ 219.712569] usb 2-6: reset SuperSpeed USB device number 2 using xhci_hcd [ 230.824786] usb 2-6: reset SuperSpeed USB device number 2 using xhci_hcd [ 241.110217] usb 2-6: reset SuperSpeed USB device number 2 using xhci_hcd [ 248.015814] usb 2-6: reset SuperSpeed USB device number 2 using xhci_hcd ### < ... iterated 953 times ... > ### latest: Nov 13 17:23:21 sid kernel: usb 2-6: reset SuperSpeed USB device number 2 using xhci_hcd #################################################################################################################### Another suspicion: ################## ### perhaps unrelated : ( I did not pull any jack at all ) Nov 13 18:04:27 sid root[10810]: ACPI event unhandled: jack/lineout LINEOUT plug Nov 13 18:04:27 sid root[10814]: ACPI event unhandled: jack/videoout VIDEOOUT plug Nov 13 18:19:34 sid root[11144]: ACPI event unhandled: jack/lineout LINEOUT unplug Nov 13 18:19:34 sid root[11147]: ACPI event unhandled: jack/videoout VIDEOOUT unplug Nov 13 18:19:47 sid root[11152]: ACPI event unhandled: jack/lineout LINEOUT plug Nov 13 18:19:47 sid root[11156]: ACPI event unhandled: jack/videoout VIDEOOUT plug ### this is now almost 1 h later: Nov 13 18:19:47 sid kernel: usb 2-6: USB disconnect, device number 2 <-- !!! Nov 13 18:19:47 sid kernel: sd 7:0:0:0: [sdd] Synchronizing SCSI cache Nov 13 18:19:47 sid kernel: sd 7:0:0:0: [sdd] Synchronize Cache(10) failed: Result: hostbyte=DID_NO_CONNECT driverbyte=DRIVER_OK ### followed by a "new" connect (which was defenitely not initiated by me) : Nov 13 18:19:48 sid kernel: usb 2-6: new SuperSpeed USB device number 3 using xhci_hcd Nov 13 18:19:48 sid kernel: usb 2-6: New USB device found, idVendor=174c, idProduct=55aa Nov 13 18:19:48 sid kernel: usb 2-6: New USB device strings: Mfr=2, Product=3, SerialNumber=1 Nov 13 18:19:48 sid kernel: usb 2-6: Product: mkn_LC-53U3-Becrux Nov 13 18:19:48 sid kernel: usb 2-6: Manufacturer: mkn_LC-POWER Nov 13 18:19:48 sid kernel: usb 2-6: SerialNumber: 201401000056 Nov 13 18:19:48 sid kernel: usb 2-6: UAS is blacklisted for this device, using usb-storage instead Nov 13 18:19:48 sid kernel: usb 2-6: UAS is blacklisted for this device, using usb-storage instead Nov 13 18:19:48 sid kernel: usb-storage 2-6:1.0: USB Mass Storage device detected Nov 13 18:19:48 sid kernel: usb-storage 2-6:1.0: Quirks match for vid 174c pid 55aa: c00000 Nov 13 18:19:48 sid kernel: scsi host8: usb-storage 2-6:1.0 Nov 13 18:19:49 sid kernel: scsi 8:0:0:0: Direct-Access ST315003 41AS 0 PQ: 0 ANSI: 6 Nov 13 18:19:49 sid kernel: sd 8:0:0:0: Attached scsi generic sg6 type 0 Nov 13 18:19:49 sid kernel: sd 8:0:0:0: [sdd] 2930277168 512-byte logical blocks: (1.50 TB/1.36 TiB) Nov 13 18:19:49 sid kernel: sd 8:0:0:0: [sdd] Write Protect is off Nov 13 18:19:49 sid kernel: sd 8:0:0:0: [sdd] Mode Sense: 43 00 00 00 Nov 13 18:19:49 sid kernel: sd 8:0:0:0: [sdd] Write cache: enabled, read cache: enabled, doesn't support DPO or FUA Nov 13 18:19:49 sid kernel: sdd: sdd1 Nov 13 18:19:49 sid kernel: sd 8:0:0:0: [sdd] Attached SCSI disk ### This correlates to the fact that under Windows 8.1 / USB 2.0, ### the disk also 'magically' disconnects after some time; ### only mechanism to "re-awake" I found so far was POWER-OFF / - ON ### ### ==> TODO: evaluate different USB 3 Link Power Management options.
Upstream : I have forwarded these findings to Hans de Goede, Linux kernel (USB 3) developer. Will keep this report updated on informations from upstream.
( CONTINUATION of Manfred Knick from comment #1 ) Mail to "upstream" ################################################# Subject: Re: [Workaround] ASM1153E : ASM 174c:55aa problems re-loaded [quirk needed] To: Hans de Goede <hdegoede@redhat.com>, Linux Kernel USB Mailing List <linux-usb@vger.kernel.org>, Mathias Nyman <mathias.nyman@intel.com> From: Manfred_Knick <Manfred.Knick@T-Online.de> Date: Tue, 17 Nov 2015 13:20:42 +0100 Hello, Hans, thank you very much for your reading and for your proposals in your quick answer! The RESUMMEE of further differential analysis with an SSD, in short: o) Neither Cable nor Power Supply are the culprit o) Quirk for this special chip ASM1153E seems necessary o) My (foreign) FW 'update' seems to be the culprit for o) producing the disconnects and o) for poor reading rates from SSD o) Original Firmware *with* quirk performs well: o) No further "reset SuperSpeed USB device number" at all o) Maximum *netto* throughput ~ 1/2 of "SuperSpeed" := 5Gb/s with SSD o) Re-awakes from spin-down (at least under linux) Kind regards again Yours sincerely Manfred P.S.: Neverhteless I will get a "really good USB 3.0 cable" anyway, re-test, and in case of differences, I will let you know. P.P.S: Can I offer you any additional information? (Just in case, I have configured CONFIG_USB_MON=y already.) ############################################################################# ### with quirks, with 'new'/foreign FW, SAMSUNG SSD 840 256GB ############################################################################# # lsusb -d 174c: Bus 002 Device 002: ID 174c:55aa ASMedia Technology Inc. ASM1051 SATA 3Gb/s bridge # usb-devices ... T: Bus=02 Lev=01 Prnt=01 Port=05 Cnt=01 Dev#= 2 Spd=5000 MxCh= 0 D: Ver= 3.00 Cls=00(>ifc ) Sub=00 Prot=00 MxPS= 9 #Cfgs= 1 P: Vendor=174c ProdID=55aa Rev=01.00 S: Manufacturer=mkn_LC-POWER S: Product=mkn_LC-53U3-Becrux S: SerialNumber=201401000056 C: #Ifs= 1 Cfg#= 1 Atr=c0 MxPwr=0mA I: If#= 0 Alt= 0 #EPs= 2 Cls=08(stor.) Sub=06 Prot=50 Driver=usb-storage ... # ddrescue -f -b 4Ki /dev/sdd /dev/zero GNU ddrescue 1.18.1 rescued: 256060 MB, errsize: 0 B, current rate: 84041 kB/s ipos: 256060 MB, errors: 0, average rate: 179 MB/s <-- poor! opos: 256060 MB, run time: 23.73 m, successful read: 0 s ago $ dmesg | grep "reset SuperSpeed USB device number" | wc -l 2426 <------------------------------------------------------------------ <--- !!! ############################################################################# ### with quirks, with original FW, SAMSUNG SSD 840 256GB ############################################################################# Nov 17 11:24:31 sid kernel: usb 2-6: new SuperSpeed USB device number 4 using xhci_hcd Nov 17 11:24:31 sid kernel: usb 2-6: New USB device found, idVendor=174c, idProduct=55aa Nov 17 11:24:31 sid kernel: usb 2-6: New USB device strings: Mfr=2, Product=3, SerialNumber=1 Nov 17 11:24:31 sid kernel: usb 2-6: Product: USB3.0 External HDD Nov 17 11:24:31 sid kernel: usb 2-6: Manufacturer: GODO Nov 17 11:24:31 sid kernel: usb 2-6: SerialNumber: 20140100004A Nov 17 11:24:31 sid kernel: usb 2-6: UAS is blacklisted for this device, using usb-storage instead Nov 17 11:24:31 sid kernel: usb 2-6: UAS is blacklisted for this device, using usb-storage instead Nov 17 11:24:31 sid kernel: usb-storage 2-6:1.0: USB Mass Storage device detected Nov 17 11:24:31 sid kernel: usb-storage 2-6:1.0: Quirks match for vid 174c pid 55aa: c00000 Nov 17 11:24:31 sid kernel: scsi host9: usb-storage 2-6:1.0 Nov 17 11:24:32 sid kernel: scsi 9:0:0:0: Direct-Access Samsung SSD 840 PRO Seri 0 PQ: 0 ANSI: 6 Nov 17 11:24:32 sid kernel: sd 9:0:0:0: Attached scsi generic sg6 type 0 Nov 17 11:24:32 sid kernel: sd 9:0:0:0: [sdd] 500118192 512-byte logical blocks: (256 GB/238 GiB) Nov 17 11:24:32 sid kernel: sd 9:0:0:0: [sdd] Write Protect is off Nov 17 11:24:32 sid kernel: sd 9:0:0:0: [sdd] Mode Sense: 43 00 00 00 Nov 17 11:24:32 sid kernel: sd 9:0:0:0: [sdd] Write cache: enabled, read cache: enabled, doesn't support DPO or FUA Nov 17 11:24:32 sid kernel: sd 9:0:0:0: [sdd] Attached SCSI disk Nov 17 11:30:01 sid cron[13650]: (root) CMD (test -x /usr/sbin/run-crons && /usr/sbin/run-crons) # lsusb -d 174c: Bus 002 Device 004: ID 174c:55aa ASMedia Technology Inc. ASM1051 SATA 3Gb/s bridge # usb-devices ... T: Bus=02 Lev=01 Prnt=01 Port=05 Cnt=01 Dev#= 4 Spd=5000 MxCh= 0 D: Ver= 3.00 Cls=00(>ifc ) Sub=00 Prot=00 MxPS= 9 #Cfgs= 1 P: Vendor=174c ProdID=55aa Rev=01.00 S: Manufacturer=GODO S: Product=USB3.0 External HDD S: SerialNumber=20140100004A C: #Ifs= 1 Cfg#= 1 Atr=c0 MxPwr=0mA I: If#= 0 Alt= 0 #EPs= 2 Cls=08(stor.) Sub=06 Prot=50 Driver=usb-storage ... ### +++ no further "reset SuperSpeed USB device number" at all +++ ### (streaming) read (dd) rate now around half of the specified 540 MB/s # ddrescue -f -b 4Ki /dev/sdd /dev/zero GNU ddrescue 1.18.1 rescued: 256060 MB, errsize: 0 B, current rate: 53239 kB/s ipos: 256060 MB, errors: 0, average rate: 252 MB/s <-- a little bit better opos: 256060 MB, run time: 16.88 m, successful read: 0 s ago Finished ############################################################################# ### with quirks, with original FW, Seagate ST315003 41AS ############################################################################# [ 7367.141587] usb 2-6: new SuperSpeed USB device number 7 using xhci_hcd [ 7367.153387] usb 2-6: New USB device found, idVendor=174c, idProduct=55aa [ 7367.153394] usb 2-6: New USB device strings: Mfr=2, Product=3, SerialNumber=1 [ 7367.153398] usb 2-6: Product: USB3.0 External HDD [ 7367.153401] usb 2-6: Manufacturer: GODO [ 7367.153404] usb 2-6: SerialNumber: 20140100004A [ 7367.154244] usb 2-6: UAS is blacklisted for this device, using usb-storage instead [ 7367.154680] usb 2-6: UAS is blacklisted for this device, using usb-storage instead [ 7367.154686] usb-storage 2-6:1.0: USB Mass Storage device detected [ 7367.154781] usb-storage 2-6:1.0: Quirks match for vid 174c pid 55aa: c00000 [ 7367.154847] scsi host12: usb-storage 2-6:1.0 [ 7368.213877] scsi 12:0:0:0: Direct-Access ST315003 41AS 0 PQ: 0 ANSI: 6 [ 7368.213976] sd 12:0:0:0: Attached scsi generic sg6 type 0 [ 7368.214147] sd 12:0:0:0: [sdd] 2930277168 512-byte logical blocks: (1.50 TB/1.36 TiB) [ 7368.214349] sd 12:0:0:0: [sdd] Write Protect is off [ 7368.214351] sd 12:0:0:0: [sdd] Mode Sense: 43 00 00 00 [ 7368.214556] sd 12:0:0:0: [sdd] Write cache: enabled, read cache: enabled, doesn't support DPO or FUA [ 7368.238443] sdd: sdd1 [ 7368.239308] sd 12:0:0:0: [sdd] Attached SCSI disk ######################################## ### Power Supply : specified for 2000 mA ### ### Primary energy consumption @ 220V : ### 31W max (spin-up), 10W (steady) ### 52VA max (spin-up), 16VA (steady) ######################################## # lsusb -d 174c: Bus 002 Device 007: ID 174c:55aa ASMedia Technology Inc. ASM1051 SATA 3Gb/s bridge sid ~ # sid ~ # sid ~ # usb-devices | grep -B 2 -A 5 "Vendor=174c" T: Bus=02 Lev=01 Prnt=01 Port=05 Cnt=01 Dev#= 7 Spd=5000 MxCh= 0 D: Ver= 3.00 Cls=00(>ifc ) Sub=00 Prot=00 MxPS= 9 #Cfgs= 1 P: Vendor=174c ProdID=55aa Rev=01.00 S: Manufacturer=GODO S: Product=USB3.0 External HDD S: SerialNumber=20140100004A C: #Ifs= 1 Cfg#= 1 Atr=c0 MxPwr=0mA I: If#= 0 Alt= 0 #EPs= 2 Cls=08(stor.) Sub=06 Prot=50 Driver=usb-storage ### After 10' (I suppose, as specified in the ASM firmware), ### the disk performs a spin-down, but (linux 4.3.0) ### +++ re-awakes (spin-up) with the next read command +++ . ### dd of the first 100 GiB: ### (streaming) read (dd) rate of 132 MB/s (specified Maximum) # ddrescue -f -b 4Ki /dev/sdd /dev/zero -s 100Gi GNU ddrescue 1.18.1 rescued: 107374 MB, errsize: 0 B, current rate: 19988 kB/s ipos: 107374 MB, errors: 0, average rate: 132 MB/s opos: 107374 MB, run time: 13.50 m, successful read: 0 s ago Finished ### +++ no further "reset SuperSpeed USB device number" at all +++
-------- Weitergeleitete Nachricht -------- Betreff: Re: [Workaround] ASM1153E : ASM 174c:55aa problems re-loaded [quirk needed] Datum: Thu, 19 Nov 2015 15:18:12 +0100 Von: Manfred_Knick <Manfred.Knick@T-Online.de> An: Hans de Goede <hdegoede@redhat.com>, Linux Kernel USB Mailing List <linux-usb@vger.kernel.org>, Mathias Nyman <mathias.nyman@intel.com> Hello, Hans, Finally I was able to _verify_ that the ASM1153E itself *is* capable of UAS under linux kernel 4.3.0 *without* quirks. Further differential analysis lead to the following conclusion: o) Fimware mess o) Fimware configuration mess o) Firmware correction / update neglectance o) Bad power supply ... all by "LC-Power". who straightforward reject any firmware update support at all. ASMedia straightforward rejects any direct delivery of firmware updates whithout end manufacturer 'testing' involvement. Luckily I found an appropriate firmware update to get the external boxes working. HTH ... Thanks! Kind regards Manfred ############################################################################# As usual, short story first: ########### o) The original firmware *claims* to supply UAS, but doesn't. o) A newer version of MPTool and firmware exists and does support UAS. After flashing the update: A) Equipped with the SSD: A.1) Using a classical stabilized laboratory power supply: o) The whole 256GB of the SSD get successfully being read o) - with almost 400 MB/s average rate (!) o) First 'read' (dd) after power-on, quite early: o) - only one (!) "ERROR Transfer event" and following "reset SuperSpeed USB device" o) Further 'read' (dd) iterations: o) - without any (!) "ERROR Transfer event" at all A.2) Using the supplied power supply: o) - 9 x "ERROR Transfer event" o) - *Failure* after ~ 10 GiB ### To me, it seems that the shortcomings of LC's Power Supply ### show up due to *quick transfer* rates ( ==> 5V ) ; ### they do not correspond to the current drawn by the motors ( ==> 12V ) , ### because: B) Equipped with the 'slow', but 'heavy' rotating Seagate: B.1) Using a classical stabilized laboratory power supply: o) - Reading 10 GiB @ 130 MB/s average without any "ERROR" B.2) Using the supplied power supply: o) - Reading 10 GiB @ 130 MB/s average without any "ERROR" C) Equipped with a WD Re 500GB ( WD5003ABYZ ): C.1) Using a classical stabilized laboratory power supply: o) - Reading 10 GiB @ 139 MB/s average without any "ERROR" C.2) Using the supplied power supply: o) - Reading 10 GiB @ 138 MB/s average without any "ERROR" D) Equipped with a WD Green 6.0TB ( WD60EZRX ): D.1) Using a classical stabilized laboratory power supply: o) - Reading 10 GiB @ 173 MB/s average without any "ERROR" D.2) Using the supplied power supply: o) - Reading 10 GiB @ 176 MB/s average without any "ERROR" ### Notabene: I'm still using LC's supplied USB 3.0 cable (could not get a really excellent replacement yet). ############################################################################# # Firmware: On the bottom of . --> http://www.computerbase.de/forum/showthread.php?t=1473501 I finally found the link . . . "Achtung: Mit UASP" . . . "Aktuelle Firmware für ASM 1053/1053E : _ASMedia.1053/1053E_ Support 8TB+ HDD + UASP" pointing to . . . http://www.computerbase.de/forum/attachment.php?attachmentid=501469&d=1435869989 Take care: Both downloads (Top: with / Bottom: without UASP) will end up into the same file name : . . . ASMedia.1153e.NoName.Device.7z Content of version "with UASP" : 60K 12. Jul 1958 140509_A1_82_00.bin <--- with 10' 60K 16. Jul 07:34 140509_A1_82_80.bin <- (without) 4,0K 2. Sep 05:38 Alternativ.MPTool.2.1.20.1 <------- ########################### Just to have it documented: Early in the progress of this investigation, I had also updated the firmware of my motherboard ASUS P9D-WS to the latest available BIOS 2202 [ 2015/07/27 update]. ############################################################################# # A) Using a classical stabilized laboratory power supply / supplied USB 3.0 cable: ... ############################################################################# # B) Using the supplied power supply / supplied USB 3.0 cable: ...
( In completion of Manfred Knick from comment #3) *RESUMEE* +++ The ASM1153E itself *is* capable of UAS +++ Especially, it was by no way any fault in the kernel: +++ linux kernel 4.3.0 USB code works perfectly *without* quirks. Unfortunately, other errors multiplied making it extremely laborious to elaborate: > o) Fimware mess > o) Fimware configuration mess > o) Firmware correction / update neglectance > o) Bad power supply ... all by "LC-Power" and, in addition: o) a bad USB 3.0 cable (Type-B ending) ... again by "LC-Power" Web research reveals that this combination is not uncommon with other manufacturers as well (so keep being suspicious). HTH ! Thanks. Manfred
> ( In completion of Manfred Knick from comment #5) *SOLUTION* - short version : If you want to exploit real 'SuperSpeed' transfers up to around ~ 400 MiB/s like cited above, you need: o) good stabilized power supply for the external adding capacitors ( 4700 micro-Farad per _maximum_ current [Ampere] drawn by your disk ) should be a very cheap solution of striking simplicity o) good highly shielded USB 3.0 cables o) up-to-date firmware ( ASM _and_ mobo ) will be essential o) up-to-date kernel will not be obstructive HTH ! Thanks. Manfred
