Stránka 1 z 1

LSI MegaRAID SAS 8704ELP - 4x pipnutie

Napsal: 23 čer 2015, 22:44
od rnbw
Dostal som nefunkcnu kartu LSI MegaRAID SAS 8704ELP. Celkom by sa hodila na testovanie SAS diskov.

Po zapnuti karta pipa (ma hnusne silne piezo) 4x, pauza, 4x, pauza... Neobjavi sa ani BIOS karty. Myslel som, ze je poskodeny firmware. Normalne to preflshovat neslo, ale cez mode 0 ano - bez zmeny. Potom som sa hral s programom megarec, "podarilo" sa mi zmazat celu flash a uz to ani nepipalo, obnova nefungovala. Tak som odspajkoval flash (Intel 28F320C3B) a naprogramoval v programatore - a uz to zase pipa.

Nikde som nenasiel vyznam tohoto chyboveho kodu. Predpokladam vsak, ze procesor zistil nejaku fatalnu chybu. Na karte su osadene 3 DRAM cipy, zial v BGA puzdrach, inak by som ich uz na skusku vymenil.

Nejake napady?

Napsal: 23 čer 2015, 23:18
od rnbw
Karta ma konektor pre seriovu konzolu - J3. Uroven je 3,3V, zapojenie: TX GND RX 3V3, parametre 115200 8N1. Po zapnuti z toho vylezie toto:

Kód: Vybrat vše

LSI Logic Boot loader
EepromInit: Family=33, SN=3fe7a0020000
Programming Product Ids 1000  1012  1


LSI Logic 1078 initialization code
Frequencies: CPU: 500 Mhz, PLB: 166 Mhz, DDR: 333 Mhz
 Strap Regs: STRP0=ff1f9505, STRP1=00000021, STRP2=2a2359f0, STRP3=f3920060
             STRP4=10000104, STRP5=00006010, STRP6=00170bfb, STRP7=00000000
   PPC ECID: 003e0000:0a076c60:0e78a80f:02401c70
>> SPD data begin <<
     Memory Type: DDR-2
    Memory Speed: 333 Mhz [3.0ns]
      Data Width: 40 bits
       #RAS Bits: 13
       #CAS Bits: 10
#Banks each chip: 4
      Chip width: x16
  #Banks in DIMM: 1
       DIMM Size: 128MB (512MB calculated from other SPD data)
     CAS Latency: CL3, CL4, CL5,
      Registered: No
Voltage Standard: SSTL 1.8
DIMM Config Type: ECC
    SPD Checksum: 0xd5 (checksum is Good)
>> SPD data end <<
RAW SPD Data:
00000000: 80 08 08 0d 0a 00 28 00 - 05 30 45 02 82 10 10 00
00000010: 0c 04 38 00 20 00 20 3d - 45 50 45 3c 28 3c 2d 20
00000020: 20 27 10 17 3c 1e 1e 00 - 00 3c 69 80 18 22 00 56
00000030: 3f 7e 24 28 38 48 48 6d - 2f 42 00 00 00 00 12 d5
tlbMapAddress: Address d0000000 (size = 10000000) overlaps TLB entry c
tlbMapAddress: Address 00000000 (size = 1000000) overlaps TLB entry 1e

SBR read successfully, 256 bytes
Extended SBR image status: 1
Raw SBR Image data:
00000000: ff 1f 95 05 00 00 00 21 - 2a 23 59 f0 f3 92 00 60
00000010: 10 00 01 04 00 00 60 10 - 00 17 0b fb 00 00 00 00
00000020: ff ff ff ff ff ff ff ff - ff ff ff ff ff ff ff ff
00000030: ff ff ff ff ff ff ff ff - ff ff ff ff ff ff ff ff
00000040: ff ff ff ff ff ff ff ff - ff ff ff ff ff ff ff ff
00000050: ff ff ff ff ff ff ff ff - ff ff ff ff ff ff ff ff
00000060: ff ff ff ff ff ff ff ff - ff ff ff ff ff ff ff ff
00000070: ff ff ff ff ff ff ff ff - ff ff ff ff ff ff ff ff
00000080: ff ff ff ff ff ff ff ff - ff ff ff ff ff ff ff ff
00000090: ff ff ff ff ff ff ff ff - ff ff ff ff ff ff ff ff
000000a0: ff ff ff ff ff ff ff ff - ff ff ff ff ff ff ff ff
000000b0: ff ff ff ff ff ff ff ff - ff ff ff ff ff ff ff ff
000000c0: ff ff ff ff ff ff ff ff - ff ff ff ff ff ff ff ff
000000d0: ff ff ff ff ff ff ff ff - ff ff ff ff ff ff ff ff
000000e0: ff ff ff ff ff ff ff ff - ff ff ff ff ff ff ff ff
000000f0: ff ff ff ff ff ff ff ff - ff ff ff ff ff ff ff ff
RFOS=3f, RDSS=2, CLKP=1, LLWP=0, WDTR=3, WDFD=0, TXDRVPWR=5, TXCA=8
Finding DDR RFDC Window [RQDC=41]: NONE
initDDR: Unable to find stable RQDC/RFDC window
!!! Firmware initialization failed. Fault code = 0x1 !!!


Takze skutocne to vyzera na problem s RAM.

Napsal: 23 čer 2015, 23:44
od rnbw
RQDC bude pravdepodobne "Read DQS Delay Control".
RFDC bude pravdepodobne "Read Feedback Delay Control".

Nasiel som to v tomto zaujimavom dokumente:
https://www-01.ibm.com/chips/techlib/te ... _fspec.pdf
Je to sice uplne nieco ine, ale su tam zaujimave veci ohladom prace s DDR pamatami - napriklad to, ze sa tieto hodnoty kalibruju automaticky.

Zapise sa hodnota do RAM a nasledne sa cita pri roznych nastaveniach registrov. Ak sa hodnota nezhoduje nikdy, tak to skonci s chybou.

Pamate su Micron D9HRC, skusim ich prepiect, pripadne vymenit (mam tu nejake vadne 512MB DDR2 moduly, malo by to sediet).

Napsal: 24 čer 2015, 22:50
od rnbw
Prepiekol som dva pamatove cipy pri procesore, ziadna zmena. Potom som prepiekol ten treti (co je tam kvoli ECC) a zacalo to fungovat :D Takze objednam kabel z eBay a mozem testovat disky.

Kód: Vybrat vše

LSI Logic Boot loader
EepromInit: Family=33, SN=3fe7a0020000
Programming Product Ids 1000  1012  1


LSI Logic 1078 initialization code
Frequencies: CPU: 500 Mhz, PLB: 166 Mhz, DDR: 333 Mhz
 Strap Regs: STRP0=ff1f9505, STRP1=00000021, STRP2=2a2359f0, STRP3=f3920060
             STRP4=10000104, STRP5=00006010, STRP6=00170bfb, STRP7=00000000
   PPC ECID: 003e0000:0a076c60:0e78a80f:02401c70
>> SPD data begin <<
     Memory Type: DDR-2
    Memory Speed: 333 Mhz [3.0ns]
      Data Width: 40 bits
       #RAS Bits: 13
       #CAS Bits: 10
#Banks each chip: 4
      Chip width: x16
  #Banks in DIMM: 1
       DIMM Size: 128MB (512MB calculated from other SPD data)
     CAS Latency: CL3, CL4, CL5,
      Registered: No
Voltage Standard: SSTL 1.8
DIMM Config Type: ECC
    SPD Checksum: 0xd5 (checksum is Good)
>> SPD data end <<
RAW SPD Data:
00000000: 80 08 08 0d 0a 00 28 00 - 05 30 45 02 82 10 10 00
00000010: 0c 04 38 00 20 00 20 3d - 45 50 45 3c 28 3c 2d 20
00000020: 20 27 10 17 3c 1e 1e 00 - 00 3c 69 80 18 22 00 56
00000030: 3f 7e 24 28 38 48 48 6d - 2f 42 00 00 00 00 12 d5
tlbMapAddress: Address d0000000 (size = 10000000) overlaps TLB entry c
tlbMapAddress: Address 00000000 (size = 1000000) overlaps TLB entry 1e

SBR read successfully, 256 bytes
Extended SBR image status: 1
Raw SBR Image data:
00000000: ff 1f 95 05 00 00 00 21 - 2a 23 59 f0 f3 92 00 60
00000010: 10 00 01 04 00 00 60 10 - 00 17 0b fb 00 00 00 00
00000020: ff ff ff ff ff ff ff ff - ff ff ff ff ff ff ff ff
00000030: ff ff ff ff ff ff ff ff - ff ff ff ff ff ff ff ff
00000040: ff ff ff ff ff ff ff ff - ff ff ff ff ff ff ff ff
00000050: ff ff ff ff ff ff ff ff - ff ff ff ff ff ff ff ff
00000060: ff ff ff ff ff ff ff ff - ff ff ff ff ff ff ff ff
00000070: ff ff ff ff ff ff ff ff - ff ff ff ff ff ff ff ff
00000080: ff ff ff ff ff ff ff ff - ff ff ff ff ff ff ff ff
00000090: ff ff ff ff ff ff ff ff - ff ff ff ff ff ff ff ff
000000a0: ff ff ff ff ff ff ff ff - ff ff ff ff ff ff ff ff
000000b0: ff ff ff ff ff ff ff ff - ff ff ff ff ff ff ff ff
000000c0: ff ff ff ff ff ff ff ff - ff ff ff ff ff ff ff ff
000000d0: ff ff ff ff ff ff ff ff - ff ff ff ff ff ff ff ff
000000e0: ff ff ff ff ff ff ff ff - ff ff ff ff ff ff ff ff
000000f0: ff ff ff ff ff ff ff ff - ff ff ff ff ff ff ff ff
RFOS=3f, RDSS=2, CLKP=1, LLWP=0, WDTR=3, WDFD=0, TXDRVPWR=5, TXCA=8
Finding DDR RFDC Window [RQDC=41]: 0x1ba - 0x2a0 [Size = 0x0e6]. Center = 0x22d
Cache was not detected dirty
Bios disable jumper not asserted
Testing memory [addr=0x0, 128 MB]: Done
Scrubbing DramPersistent [addr=0x80000000-0x80040000]
Scrubbing between DramPersistent and TTY [addr=0x80040000-0x80100000]

loadBios: Images found=2 start=80040000 size=10800
Scrubbing TTY [addr=0x80100000-0x80200000]
Scrubbing from end of TTY to Disk Cache [addr=0x80200000-0x81600000]
Scrubbing Disk Cache [addr=0x81600000, 106 MB]

**** Memory Test Options [3 seconds]: 'a' = Test, 'b' = Calibrate:
LOAD section: src=fe00ad8e, size=ab9a1, dst=80c00000, mode=1...done
LOAD section: src=fe0b6738, size=26b4, dst=80d500e0, mode=1...done
LOAD section: src=fe0b8df5, size=3ed, dst=e401e400, mode=1...done

T0: LSI Logic ROC firmware
T0: Copyright (C) LSI Logic, 2004
T0: Firmware version 1.40.342-1650 built on May 11 2012 at 11:47:31

T0: pci=60 sbr=60
T0: pciInit: O_PCI_SERVICE = 00000001
T0: Initializing memory pool size=002F3B24 bytes
T0: Press '!' within 3 seconds to enter debugger before INIT
T3: Headless System 0
T3: EVT#117694-T3:   0=Firmware initialization started (PCI ID 0060/1000/1012/1000)
T3: EVT#117695-T3:   1=Firmware version 1.40.342-1650
T3: EepromInit: Family=33, SN=3fe7a0020000
T3: Authenticating RAID key: Done!
T3: I2Chandle obtained for MUX [0]0x0
T3: I2Chandle obtained for MUX [1]0x10
T3: DRAM SIZE=128 MB
T3: Setting Connector Defaults
T3: Setting PHY defaults

T3: Environment data:
T3: MAXFLUSHCMD=10000096
T3: VERSIONS=APP_1.40.342-1650,BIOS_2.07.00,BCON_2.2-22-e_15-Rel,PCLI_01.40-010:#%00008,BTBL_1.00.00.01-0012
T3: PACKAGE=11.0.1-0048
T3: VALIDATION=GCA_05/23/12_00:52:27

T3: MFC data:
T3:     subVendorId/subDeviceId=1000/1012
T3:     OEM=1, clusterDisable=1, disableSAS=0, maxDisks=1, enableRaid6=1, disableWideCache=0
T3:     disableRaid5=0 enableSecurity=0
T3:     sasAddr=500605b000c5c530, phyPolarity=00 phyPolaritySplit f0
T3:     backgroundRate=30, stripeSize=7(64K), flushTime=4
T3:     writeBack=1, readAhead=0, cacheWhenBBUBad=0, cachedIo=0
T3:     smartMode=0(6), alarmDisable=0, coercion=2(1G), zcrConfig=0(Undefined)
T3:     dirtyLedShowsDriveActivity=0, biosContinueOnError=0, spindownMode=0(None)
T3:     allowedDeviceTypes=0(SAS/SATA), allowMixInEnclosure=1, allowMixInLD=1, allowSataInCluster=0
T3:     allowSSDMixInLD=0, allowMixSSDHDDInLD=0
T3:     maxChainedEnclosures=3, disableCtrlR=1, enableWebBios=1, directPdMapping=0, biosEnumerateLds=1
T3:     restoreHotSpareOnInsertion=0, exposeEnclosureDevices=1
T3:     maintainPdFailHistory=1 disablePuncturing=0 zeroBasedEnclEnumeration=0 disableBootCLI=0
T3:     quadPortConnectorMap=0 driveActivityLed=0 disableAutoDetectBackplane=0 enableLedHeaders=0 useFdeOnly=0 delayPOST=0
T3: MFC MFG data:
T3:     date=08/15/08, sn="P019263308", reworkDate=00/00/00, rev=""
T3: No PHY_GB command was found in environment.  So using default settings.

T3: can_flush = 0

T3: Controller Properties: (All values are in hex)
T3:    seqNum=1, predFailPollInterval=12c, intThrottleCount=10
T3:    intThrottleTimeUs=32, rebuildRate=1e, patrolReadRate=1e
T3:    bgiRate=1e, ccRate=1e, reconRate=1e
T3:    cacheFlushInterval=4, spinupDriveCount=2, spinupDelay=c
T3:    clusterEnable=0, coercionMode=2, alarmEnable=1
T3:    disableAutoRebuild=0, disableBatteryWarning=1, eccBucketSize=f
T3:    eccBucketLeakRate=5a0, restoreHotSpareOnInsertion=0
T3:    exposeEnclosureDevices=1, maintainPdFailHistory=1
T3:    disallowHostRequestReordering=0
T3:    abortCCOnError=0, prCorrectUnconfiguredAreas=0
T3:    copyBackDisabled=0, SMARTerEnabled=0
T3:    enableSpinDownUnconfigured=0, LoadBalanceEnabled=1  disableNCQ=0
T3: disableOnlineCtrlReset=0
T3: OnOffProperties=800bf840, I2Chandle obtained for Ni-MH based BBU Charger 1
T3: I2Chandle obtained for Gas Guage 2
T3: I2Chandle obtained for Li-Ion based BBU Charger 3
T3: I2Chandle obtained for Li-Ion based BBU I2C Buffer 4
T3: I2Chandle obtained for I2C_SEP 5
T3: Quantity of unqualified learn cycles = 0
T3: No Battery is Present
T3: PdInit: PD_INFO entries are 240 bytes
T3: LdInit: LD_INFO entryes are 200 bytes
T3: EVT#117696-T3: 266=Board Revision
T3: ready_list=81600000 (20000 elements)
T3: rebuild_list=81700000 (8400 elements)
T3: lines_info=81742000 (28400 elements)
T3: DISK_CACHE_ADDR=819c6000
T3: MEM_END_ADDR=87fffff0
T4: KM_KeyMgmtInitTRNG: security is not enabled
T4:  Device Module Version 1.1 Date May 11 2012 Time 11:47:12
T4: Link discovery delay 10 Seconds
T4: DM_ChipInit: Total chips=1
T4: Chip=0 vendorId 1000 deviceId 0060
T4: Pl firmware version 1.25.172.0
T4: PL MID #0: msgNm=00000000, addr=0159c900
T4:  SgInIo Frame 11 SgInSg Frame 15 TotSg Frames 7fe Frame Size 100
T4: DM_PL_FillSasIoUnitPage1 : sataMaxQDepth=8
T4: iopiConfig(): Unhandled Request (PageType=9, PageNum=7)
T4: NCQ Mode value is not valid or not found, return default
T4: NCQ Mode value is not valid or not found, return default
T4:  Page8 being configured for Vitesse SEP
T4: NCQ Mode value is not valid or not found, return default
T4: Waiting for First PL Discovery to complete
T5: MPI_EVENT_SAS_DISCOVERY: PortBitmap ff - Discovery is in progress
T5:  Unsupported Event 10010000, pEventReply=15d3a00
T5: MPI_EVENT_SAS_DISCOVERY: PortBitmap 0 - Discovery is complete
T5: Disc-prog= 0....resetProg=0 aenCount=0 transit=0
T6: MPI_EVENT_SAS_DISCOVERY: PortBitmap ff - Discovery is in progress
T6: MPI_EVENT_SAS_DISCOVERY: PortBitmap 0 - Discovery is complete
T6: Disc-prog= 0....resetProg=0 aenCount=0 transit=0
T17: First discovery completed in 12 seconds
T17: MPT_SASIoUnitGetCurrentPage 0 :SASIoUnit = 815dcd40 size 90
T17: Hdr: ExtPageLength 0, ExtPageType 0 PageNumber 0
T17: PageType 0 PageVersion 0 NumPhys 8
T17: PHY0: Port=0, PortFlags=1, PhyFlags=0, NegRate0, CtrlPhyDevInf=71 AttDevHdl=0 CtrlDevHdl=1 DiscSts=0
T17: PHY1: Port=1, PortFlags=1, PhyFlags=0, NegRate0, CtrlPhyDevInf=71 AttDevHdl=0 CtrlDevHdl=1 DiscSts=0
T17: PHY2: Port=2, PortFlags=1, PhyFlags=0, NegRate0, CtrlPhyDevInf=71 AttDevHdl=0 CtrlDevHdl=1 DiscSts=0
T17: PHY3: Port=3, PortFlags=1, PhyFlags=0, NegRate0, CtrlPhyDevInf=71 AttDevHdl=0 CtrlDevHdl=1 DiscSts=0
T17: PHY4: Port=4, PortFlags=1, PhyFlags=6, NegRate1, CtrlPhyDevInf=71 AttDevHdl=0 CtrlDevHdl=5 DiscSts=0
T17: PHY5: Port=5, PortFlags=1, PhyFlags=6, NegRate1, CtrlPhyDevInf=71 AttDevHdl=0 CtrlDevHdl=5 DiscSts=0
T17: PHY6: Port=6, PortFlags=1, PhyFlags=6, NegRate1, CtrlPhyDevInf=71 AttDevHdl=0 CtrlDevHdl=5 DiscSts=0
T17: PHY7: Port=7, PortFlags=1, PhyFlags=6, NegRate1, CtrlPhyDevInf=71 AttDevHdl=0 CtrlDevHdl=5 DiscSts=0
T17: DISM: Queued!
T17: DM_DevMgrIsChipInit 0 State 400
T17:  DM_DevNotifyRAID: Notify Done. Check for Removal
T17:  DISM Complete
T17: gDevInfo=81217920, size=140
T17: Total Device = 0 
T17: PhyId 0 Sas 0 Type 0 IsSata 0, Smp 0:0
T17: PhyId 0 Sas 0 Type 0 IsSata 0, Smp 0:0
T17: PhyId 0 Sas 0 Type 0 IsSata 0, Smp 0:0
T17: PhyId 0 Sas 0 Type 0 IsSata 0, Smp 0:0
T17: PhyId 0 Sas 0 Type 0 IsSata 0, Smp 0:0
T17: PhyId 0 Sas 0 Type 0 IsSata 0, Smp 0:0
T17: PhyId 0 Sas 0 Type 0 IsSata 0, Smp 0:0
T17: PhyId 0 Sas 0 Type 0 IsSata 0, Smp 0:0
T17: Load Balance Statistics Path0PDs 0 Path1PDs 0
T17:  DM_DevMgr become Ready- state 11
T17: EnclSgpioAdd: find ICO_SGPIO type Numslot=8, EnclHandle=1, StartSlot=0
T17: Frames start at 80400000
T17: KM_KeyMgmtInit: security is not enabled
T17: The Most Recent & Current Configuration is missing
T17: EVT#117697-T17: 241=Previous configuration completely missing at boot
T40: pdFailHistInit: count=1
T40: 3a07719e
T41: pdFailHistInit: count=1
T41: 3a07719e
T41: Initializing write journal entries
T41: NOT Flushing Cache
T41: RMW: NVRAM structure valid - checking for active RMWs
06/24/15 21:51:56: EVT#117698-06/24/15 21:51:56:  44=Time established as 06/24/15 21:51:56; (46 seconds since power on)

Napsal: 24 čer 2015, 23:02
od forbidden
Dobrý, to stačilo fakt jen propíct, nebo jsi to překuličkoval? Nám tu konečně po letech dovolili koupit novej server, taky jsem si tam nakonfiguroval SAS disky, ale mimo ten server je taky není jak a kam připojit. Naštěstí máme teď 3 roky NBD, tak nemusím mít obavy. :)

Napsal: 24 čer 2015, 23:12
od rnbw
Iba som to teplovzdusnou pistolou prepiekol.

My mame nove aj stare servery. Do tych starych obcas davame pouzite disky, takze je dobre ich predtym otestovat - pozriet SMART, urobit self testy a pripadne upgradovat firmware.

Donedavna som na testovanie pouzival jeden server, co bol navyse, ale ten uz je v prevadzke. Naposledy som musel testovat disky v produkcnom serveri, kde su volne pozicie...nicomu to nevadi, ale je to dost nepohodlne, lebo tam musim ist autom.