PR ID/CR ID:	LSIP200044516
Headline:	Add mfc paramater that enables prints of topology changes in fw serial port output
Description:	Gen2 fw disables prints of topology changes in fw serial port output. Prints are temporarily redirected to tty log. Test engineers, developers, customers, can later obtain the prints by dumping the tty log. This is inconvenient for testers. In particular, this OEM uses the text in the fw logs during their automated tests of our MegaRAID products. They use scripts that look for topology changes such as drive state changes. They created the scripts based on our gen1 products, which do not supress the prints in fw. Now in fw for Gen2 products (SS, AF, 10M01, etc.) the prints are supressed for topology changes. The customer was initially misinterpretting the missing drive state changes as MegaRAID failures. Eventually, we discovered it was because of missing text in fw logs. The customer cannot run their automated tests. The customer requests an mfc or similar paramater that will allow them to enable fw log prints of topology changes.
State:		Implementation

PR ID/CR ID:	LSIP200086240
Headline:	FW should issue STANDBY IMMEDIATE command to SATA disks during shutdown from OS
Description:	1) One line description of the customer issue (customer pain point). Request a power off sequence that is less harmful to the drive during normal and emergency shutdown from OS. 2) List of affected components. Indicate pertinent HW, pre-boot apps, CLI, UX, etc. FW 3) Describe the current customer use cases. How is the product used today by this customer? What problems are encountered? Attach any pertinent logs and supporting information. Section 12.3 in the attached spec from the drive vendor Hitachi mentions ... Problems can occur on most HDDs when power is removed at an arbitrary time. Examples: 1. Data loss from the write buffer. 2. If the drive is writing a sector, a partially-written sector with an incorrect ECC block results. The sector contents are destroyed and reading that sector results in a hard error. 3. Heads possibly land in the data zone instead of landing zone, depending on the design of the HDD. FTS mentioned that during investigation of problems around command timeout / medium errors of the new BC-SATA HDD generations, it turned out that during normal shutdown (& emergency shutdown) of OS, no STBY IMMEDIATE command is sent to the drives. This leads in all cases to emergency unloads of the HDD, which are 100x more stressful as a normal unload and is shorten the HDD lifetime. 4) Describe the desired customer use cases. What behavior does the customer wish to see? How, specifically, does it differ from current behavior? What components does the customer request to be changed? Will any desired customization be under customer control, or remain under LSI control? You may then turn off the HDD in the following order: 1.Issue Standby Immediate or sleep command. 2.Wait until COMMAND COMPLETE STATUS is returned. (Drive manufacturer Hitachi mentions that it may take up to 350 ms in typical case. FTS mentions that in a typical case 500 ms are required for the command to finish completion; however, the host system time out value needs to be 30 seconds considering error recovery time) 3.Terminate power to HDD. This power-down sequence should be followed for entry into any system power-down state, or system suspend state, or system hibernation state. In a robustly designed system, emergency unload is limited to rare scenarios such as battery removal during operation. 5) Does OEM require customization (e.g., customized behavior)? State the level of customization (i.e., does the level of control remain w/ OEM (e.g., NVDATA) and/or w/ END-USER (e.g., controller properties)?) 6) Urgency. Describe the relative severity of the problem, in the customer‚s view. Are there time-critical considerations to be aware of? 7) Other supporting information (e.g., system vendor & model, OS and version, component revision levels, version of DAS component.) Drive vendor: Hitachi see section 12.3 in the attached Hitachi spec sheet
State:		Implementation

IMP ID/CO ID:	LSIP200164222
Headline:	(auto4comp FW_SAS_LSI) Add mfc paramater that enables prints of topology changes in fw serial port output
Description:	Auto Create from PR LSIP200044516. Gen2 fw disables prints of topology changes in fw serial port output. Prints are temporarily redirected to tty log. Test engineers, developers, customers, can later obtain the prints by dumping the tty log. This is inconvenient for testers. In particular, this OEM uses the text in the fw logs during their automated tests of our MegaRAID products. They use scripts that look for topology changes such as drive state changes. They created the scripts based on our gen1 products, which do not supress the prints in fw. Now in fw for Gen2 products (SS, AF, 10M01, etc.) the prints are supressed for topology changes. The customer was initially misinterpretting the missing drive state changes as MegaRAID failures. Eventually, we discovered it was because of missing text in fw logs. The customer cannot run their automated tests. The customer requests an mfc or similar paramater that will allow them to enable fw log prints of topology changes.
State:	Delivered
Associated Task:	LSIP200060486
Task References:	LSIP200044516

IMP ID/CO ID:	LSIP200167677
Headline:	(auto4comp FW_SAS_LSI) FW should issue STANDBY IMMEDIATE command to SATA disks during shutdown from OS
Description:	Auto Create from PR LSIP200086240. 1) One line description of the customer issue (customer pain point). Request a power off sequence that is less harmful to the drive during normal and emergency shutdown from OS. 2) List of affected components. Indicate pertinent HW, pre-boot apps, CLI, UX, etc. FW 3) Describe the current customer use cases. How is the product used today by this customer? What problems are encountered? Attach any pertinent logs and supporting information. Section 12.3 in the attached spec from the drive vendor Hitachi mentions ... Problems can occur on most HDDs when power is removed at an arbitrary time. Examples: 1. Data loss from the write buffer. 2. If the drive is writing a sector, a partially-written sector with an incorrect ECC block results. The sector contents are destroyed and reading that sector results in a hard error. 3. Heads possibly land in the data zone instead of landing zone, depending on the design of the HDD. FTS mentioned that during investigation of problems around command timeout / medium errors of the new BC-SATA HDD generations, it turned out that during normal shutdown (& emergency shutdown) of OS, no STBY IMMEDIATE command is sent to the drives. This leads in all cases to emergency unloads of the HDD, which are 100x more stressful as a normal unload and is shorten the HDD lifetime. 4) Describe the desired customer use cases. What behavior does the customer wish to see? How, specifically, does it differ from current behavior? What components does the customer request to be changed? Will any desired customization be under customer control, or remain under LSI control? You may then turn off the HDD in the following order: 1.Issue Standby Immediate or sleep command. 2.Wait until COMMAND COMPLETE STATUS is returned. (Drive manufacturer Hitachi mentions that it may take up to 350 ms in typical case. FTS mentions that in a typical case 500 ms are required for the command to finish completion; however, the host system time out value needs to be 30 seconds considering error recovery time) 3.Terminate power to HDD. This power-down sequence should be followed for entry into any system power-down state, or system suspend state, or system hibernation state. In a robustly designed system, emergency unload is limited to rare scenarios such as battery removal during operation. 5) Does OEM require customization (e.g., customized behavior)? State the level of customization (i.e., does the level of control remain w/ OEM (e.g., NVDATA) and/or w/ END-USER (e.g., controller properties)?) 6) Urgency. Describe the relative severity of the problem, in the customer‚s view. Are there time-critical considerations to be aware of? 7) Other supporting information (e.g., system vendor & model, OS and version, component revision levels, version of DAS component.) Drive vendor: Hitachi see section 12.3 in the attached Hitachi spec sheet
State:	Delivered
Associated Task:	LSIP200060487
Task References:	LSIP200086240

DFCT ID:	LSIP200157150
Customer DFCT No:	HSAXXXX
Headline:	(CL LSIP200140930) (HSAXXXX) The new event can't be logged.
Description:	The new event can not be logged and user also can't dump the event. The FW does not recognize the end of the event log region in flash. The problem is when FW reaches the end of event log flah region, it wraps around and start using the beginning of the region (overwrite the oldest event log). The flash region for the event log is 0xfe600000 to 0xfe760000. However, before 10M03, it used to be 0xfe600000 to 0xfe7e0000. FW started using the end of 0x80000 for "ttyloginFlash". FW starts writing the tty log from 0xfe760000 and wrap the event log at 0xfe75ffff that is OK. However it still tries to read the event data beyond 0xfe75ffff (till 0xfe7e0000).
Steps to Reproduce:	1. log event till see that "the event log wrapped" 2. dump the event log.
Resolution:	This issue introduced when we added "ttyLogInFlash" feature in the FW. FW uses 512KB of flash region for MR that is located in the end of event log section in the flash. Since FW added the "ttyLogInflash" feature, we have to change the event log size only for MR. Also,the event log size was declared as 2MB for iMR which is wrong. The correct size is 1MB.
Customer Defect Track No:	HSAXXXX
Fix Impact:	Low
External Behavior Change:	No
Suggested Testing:	Create a lot of events. Make sure you see the "event wrapped" event. After that, make sure FW can logged the new events followed by the "wrapped" event. Also make sure you can dump all events.
Child Tasks:	LSIP200060794

DFCT ID:	LSIP200125226
Headline:	Red LED is lit on the next empty slot in an AIC/Xtor JBOD if the empy slot is immediatly after an array member
Description:	Bug reported on 8780 & 9280 only fixed on 8780, still issue with Post 10M09 GCA Patch Release 4.6 DFCT ID: LSID100126313 Headline: Red LED is lit on the next empty slot in an AIC/Xtor JBOD if the empy slot is immediatly after an array member Description: I have reproduced this on the 8880 with FW 1.40.42-0615 as well as the 9280-8E with FW 2.0.03-0635 under FC6 32bit Linux, 4.09 driver. This is NOT reproducible on the 8880 with FW 1.40.02-0514. The JBOD is an AIC/Xtor 3U 16 bay with the latest FW 1.14H.08, drives used were HDS SATA 3GB. I have recreated this numerous ways. Simplest steps to reproduce; 1. Insert three SATA disks in the array, slots 1-3. 2. MegaCli cfgclr a0 3. MegaCli cfgallfreedrv r5 a0 4. RED LED on slot 4 will be on. Tested R0/5 3,4,7 & 12 drives. Can also create a 1 drv R0 on slot 12 and slot 13 light Red will go on. Creating a 1 drv R0 on a slot with a drive present in the next slot does NOT reproduce the issue. Debug output is attached but I did not see any indication of an issue. This was originally reported to me by an AIC Eng. Manager whom I was working on a separate issue, his contact issue is below. He claims to have multiple customers reporting this issue and claims some are in stop ship because of the false error condition that it presents. This was originally reported with SAS & SATA drives on the 8888. Version of Bug Reported: 1.40.42-0615 / 2.0.03-0635 Version of Bug Fixed: 1.40.72-0702_PL-Ver-1.25.77.0 Steps to Reproduce: I have recreated this numerous ways. Simplest steps to reproduce; 1. Insert three SATA disks in the array, slots 1-3. 2. MegaCli cfgclr a0 3. MegaCli cfgallfreedrv r5 a0 4. RED LED on slot 4 will be on. Tested R0/5 3,4,7 & 12 drives. Can also create a 1 drv R0 on slot 12 and slot 13 light Red will go on. Creating a 1 drv R0 on a slot with a drive present in the next slot does NOT reproduce the issue. Resolution: Fixed Resolution Description: The bayNumber was retrieved wrongly when read from the additional element status descriptor page for the case of EIP bit is set to ‘1‚. The bayNumber is calculated correctly (at line 599) of file mapSes.c and the total disk device counter was used to do the calculation. Customer List: LSI -- LSI Fix Impact: Medium Suggested Testing: Please test with AIC/XTOR Jbod encl Child Tasks: LSID100152880
Steps to Reproduce:	See Description
Resolution:	When the EIP bit is set to '0' for enclosure status page, the bayNumber for the slot (obtained from enclosure status page) needs to be used used instead of the soft state slot index.
Suggested Testing:	Please test with an enclosure, e.g. AIC/XTOR enclosure where the EIP bit is set for enclosure status page.
Child Tasks:	LSIP200060484

DFCT ID:	LSIP200146484
Headline:	MR FW reports "Unexpected sense, Invalid field in CDB" for LOG SENSE command sent to Seagate drives
Description:	Customer reported that MR FW 10M09 (FW2.120.03-1073) always show “Unexpected sense: PD ….., CDB = 0x4d …..” on all ports after reboot and launched MSM. This symptom only happens with Seagate SAS HDD and did not happen when they use same system configuration with MR 10M06 FW (2.90.03-0928). This doesn't look fatal error but customer concerns that this will create support calls since users see many of these errors always in MSM when the system is started. Also this didn't happen with 10M06 FW.
Steps to Reproduce:	Windows Server 2008 R2 Enterprise x64 X8DTU-6TF+ Rev2.01, system BIOS 11/22/2010 Rev2.0c 1x Intel Xeon X5560 2.80GHz (Nehalem 2.80GHz-6.40GT-1333) Memory: 2GB 2Rx8 PC3-8500R-7-00-BP DDR3 1066 CL7 ECC REG (MT18JSF25672PDY-1G1D1BA) x 3 Onboard SAS2108 B4 stepping driver version: 4.32.0.64 MSM version: 8.16.0100 Drives tested Seagate ST3450802SS Seagate ST31000640SS Seagate ST3300656SS 1) Connect Seagate SAS HDD, reboot the OS and launch MSM. 2) Connect HDD with SMART or Predictive Failure, open WebBIOS or MSM to check Predictive Failure.
Resolution:	In 10M09, the feature to monitor the Drive temperature is introduced. As per this feature, the FW issues LOG_SENSE command with LOG_SENSE_INFORMATIONAL_EXCEPTIONS_PAGE page code(0x2F), for getting both Temperature & Predictive Failure Information from the Drives. If these commands fail to the drives, the drive temperature is set to 0xFF(PD_INVALID) in PdInfo and from that cycle onwards, this command will never be issued to the drive again, and fw issues only REQUEST-SENSE to monitor Predictive failure Information. When this LOG-SENSE command fails to the drives, the event, "Unexpected sense, Invalid field in CDB" is getting generated. Fix: RDM_FLAGS_FILTER_SENSE has been set for LOG-SENSE command to prevent Event generation for LOG_SENSE command failure, as this command is an internal command, and failure need not have to be reported to upper layers.
Fix Impact:	Cosmetic
External Behavior Change:	No
Suggested Testing:	With new FW, the above event will not be generated.
Child Tasks:	LSIP200060792

DFCT ID:	LSIP200166039
Headline:	Label of Virtual Disk not provided by Drive Identify Command
Description:	MegaRAID firmware does not support a mechanism of sending/storing the user defined label of a Virtual Drive Customer uses 3 RAID ctrls in one system. On each RAID ctrl the customer uses multiple Disk groups and multiple virtual drives on each disk group. To differentiate between the virtual drives the LINUX command lsscsi provides following output for a competitor product. lsscsi : [0:0:1:0] disk Adaptec VD_n V1.0 /dev/sda VD_n is the label of the generated virtual drive which has been entered during the creation of the virtual drive. Using the same command with MegaRAID and latest released SW package we do not output the name of the virtual drive, we generate a output using the default name of the MegaRAID controller: [0:2:0:0] disk LSI MR9280-24i4e 2.11 /dev/sda With MegaRAID the customer has no way to differentiate between the virtual drives. The customer plans to use 3 highport controllers 9280-24i4e / system and multiple virtual drives / controller. Handling >10 virtual drives without the corresponding label of the virtual drive is an issue for their manufacturing line. The customer plans to use 3 highport controllers 9280-24i4e / system and multiple volumes / controller. Entering lsscsi should provide the Label of each volume which requires the firmware to support a mechanism of sending/storing the user defined label of a RAID volume. Examples: IOMeter does not display the Label Names, for 10 different virtual drives on different disk groups (RAID5, RAID0 etc) we display 10 times exactly the same virtual drive label ! Setting up VDs in Filesystems: If you want to format a Virtual Disk, we do not provide any kind of info on the virtual disk, just the capacity is visible to the user. 10M09 is the latest released package not supporting this request.
Steps to Reproduce:	Customer uses 3 RAID ctrls in one system. On each RAID ctrl the customer uses multiple Disk groups and multiple virtual drives on each disk group. To differentiate between the virtual drives the LINUX command lsscsi provides following output for a competitor product. lsscsi : [0:0:1:0] disk Adaptec VD_n V1.0 /dev/sda VD_n is the label of the generated virtual drive which has been entered during the creation of the virtual drive. Using the same command with MegaRAID and latest released SW package we do not output the name of the virtual drive, we generate a output using the default name of the MegaRAID controller: [0:2:0:0] disk LSI MR9280-24i4e 2.11 /dev/sda With MegaRAID the customer has no way to differentiate between the virtual drives. The customer plans to use 3 highport controllers 9280-24i4e / system and multiple virtual drives / controller. Handling >10 virtual drives without the corresponding label of the virtual drive is an issue for their manufacturing line.
Resolution:	F/w uses the controller product identification information when populating the LD level inq data. To fix this requirement, f/w needs to copy the LD "name" when filling up LD product identification data. This change is now made for LSI controller only and also needs to be enabled by "stamping" the environmental variable "LDNAMEINLDINQ" in the f/w. The stamp value should be non-zero positive.
Suggested Testing:	Please create raid LD and give some name, from the MR management tools and see from the OS device manager whether the given name gets reflected.
Child Tasks:	LSIP200060781

DFCT ID:	LSIP200146481
Headline:	SMART predictive failure detection takes longer in 10M09 than 10M06
Description:	Customer reported that 10M09 (FW2.120.03-1073) seems to take longer than 5 minute to detect HDD SMART or Predictive Failure. They also indicates that SMART failure with some HDD is detected by 10M06 (FW2.90.03-0928) but could not be detected by 10M09.
Steps to Reproduce:	Windows Server 2008 R2 Enterprise x64 X8DTU-6TF+ Rev2.01, system BIOS 11/22/2010 Rev2.0c 1x Intel Xeon X5560 2.80GHz (Nehalem 2.80GHz-6.40GT-1333) Memory: 2GB 2Rx8 PC3-8500R-7-00-BP DDR3 1066 CL7 ECC REG (MT18JSF25672PDY-1G1D1BA) x 3 Onboard SAS2108     B4 stepping driver version: 4.32.0.64 MSM version: 8.16.0100 Drives tested Seagate     ST3450802SS Seagate     ST31000640SS Seagate     ST3300656SS 1) Connect Seagate SAS HDD, reboot the OS and launch MSM. 2) Connect HDD with SMART or Predictive Failure, open WebBIOS or MSM to check Predictive Failure.
Resolution:	Temp monitoring functionality has been isolated from the Smart Polling Logic. Now, smartPollissue fires REQUEST_SENSE to the drive and in the callbackfunction, the new function smartPdTempMonitor() is invoked. This new function issues, ATA_CMD_READ_LOG_EXT( STP command) to the SATA drives for getting the temperature information, and issues LOG-SENSE command for SAS drives for the same . If these commands fail to the drives, the drive temperature is set to 0xFF(PD_INVALID) in PdInfo and from that cycle onwards, this command will never be issued to the drive again. RDM_FLAGS_FILTER_SENSE has been set for LOG-SENSE command to prevent Event generation for LOG_SENSE command failure, as this command is an internal command, and failure need not have to be reported to upper layers.
Fix Impact:	Low
External Behavior Change:	No
Suggested Testing:	Confirm "predictive failure" event generation with SATA drives, with the new Firmware.
Child Tasks:	LSIP200060792

DFCT ID:	LSIP200153454
Customer DFCT No:	7021223
Headline:	Firmware Incompleted Writes During Power Fail and Recovery
Description:	This was created in an Exadata Storage Node (Lynx 2U) Configuration: Lynx-2U Niwot HBA 12 disk drives each a RAID 0 Volume RAID Cache in Write-back mode RHEL 5.5 OS 1. System is stable state and open writes pending 2. Power-cut for the server, and the server copletes unexpected shutdown. 3. Server power on 4. Expected to see uncompleted writes that are saved in HBA cache to be opened and completed in an orderly maner upon power back. However, this does not happen and the open and incomplete writes are missing from the log. We have identified that on every cell where we have these problems, we see the following in the event log. Event Description: Controller cache discarded due to memory/battery problems I am attaching mevt1.out, mfw1.out, mbbu1.out. madp1.out, mcfg1.out.
Steps to Reproduce:	1. System is stable state and open writes pending 2. Power-cut for the server, and the server copletes unexpected shutdown. 3. Server power on 4. Expected to see uncompleted writes that are saved in HBA cache to be opened and completed in an orderly maner upon power back. However, this does not happen and the open and incomplete writes are missing from the log.
Resolution:	If during memory initialization correctable ECC errors are detected, MRFW performs re-write of cache data. If following re-write, ECC errors were detected and these were of un-correctable type, ECC errors were ignored (treated as if correctable). In such case, cache data should be discarded and event be generated.
Customer Defect Track No:	7021223
Fix Impact:	Medium
External Behavior Change:	Yes
Suggested Testing:	This is very difficult to reproduce since reproduction involves correctable ECC errors followed by uncorrectable ECC errors.
Child Tasks:	LSIP200060818

DFCT ID:	LSIP200164210
Customer DFCT No:	DF460039
Headline:	(DF460039) Enable/disable of HDD cache doesn't appear to work
Description:	If you use the WebBIOS or MegaCLI to enable/disable the HDD cache on SAS HDDs, it does not appear to work. The cache is not enabled/disabled and a SAS trace shows no Mode Select commands are issued by the card. MegaCLI reports "success" in setting the feature
Steps to Reproduce:	Create a RAID array using the 9260 (RAID 10 in my case) with SAS HDDs Enable the HDD cache using MegaCLI. MegaCLI reports success. Run benchmark to determine perf. Disable the HDD cache using MegaCLI. MegaCLI reports success. Run benchmark to determine perf. Notice no difference in perf. Following are the actual megacli commands used. -- ./MegaCli -LDSetProp -EnDskCache -lall -aall Set Disk Cache Policy to Enabled on Adapter 0, VD 0 (target id: 0) success Exit Code: 0x00 ./MegaCli -LDGetProp -DskCache -lall -aall Adapter 0-VD 0(target id: 0): Disk Write Cache : Enabled Exit Code: 0x00 Additional config: During the competitive analysis for the customer, we‚ve encountered to the case where we need to control the ‘cache in the PDs‚. As below, the command to control (enable/disable) the cache in the PDs seems working. -- ./MegaCli -LDSetProp -EnDskCache -lall -aall Set Disk Cache Policy to Enabled on Adapter 0, VD 0 (target id: 0) success Exit Code: 0x00 ./MegaCli -LDGetProp -DskCache -lall -aall Adapter 0-VD 0(target id: 0): Disk Write Cache : Enabled Exit Code: 0x00 -- However, when we capture the actual command sequences with SAS analyzer, it turned out that the actual command - mode select with page8 WCD - is not being issued to the drive and only the mode sense is being issued. Is this a intended behavior? Or, is there a certain version of megacli that actually to control the cache in the PDs? -- MegaCLI SAS RAID Management Tool Ver 8.00.40 Oct 12, 2010 LSI FW: 12.12.0-0045 LSI Driver for VMware ESX 4.1: rpm_vmware-esx-drivers-scsi-megaraid-sas_400.4.0.14.1-18vmw.0.0.260247@x86_64
Customer Defect Track No:	DF460039
Child Tasks:	LSIP200060795

DFCT ID:	LSIP200166899
Headline:	HWR: Fw log should indicate if a created/existing VD is already initialized
Description:	Reference original PR LSIP200084297 This is a more convenient solution approved by the customer. During our automated tests we need to check electronically if created/existing VD is already initialized. MegaCLI‚s –LdInfo command is returning the state of the disk (optimal, degraded, offline) but this tells us nothing whether the disk is initialized. Especially in spanned arrays it is important, where the single spans can have different states in regards to initialized flag
Steps to Reproduce:	Create or connect any LD to a Liberator MR Try to obtain LD initialization state with MegaCLI or in megamon commands. MegaCLI‚s –LdInfo command is returning the state of the disk (optimal, degraded, offline) but this tells us nothing whether the disk is initialized.
Child Tasks:	LSIP200060796

DFCT ID:	LSIP200156699
Customer DFCT No:	CSCti85988
Headline:	MegaCli -AdpAllInfo shows BBU is not supported for Castor cards
Description:	MegaCli -AdpAllInfo shows BBU is not supported for Castor cards. This is a Castor with a BBU on it in an Alpine system.
Steps to Reproduce:	10M09 components MegaCli -AdpAllInfo -a0 Supported Adapter Operations ================ Rebuild Rate : Yes CC Rate : Yes BGI Rate : Yes Reconstruct Rate : Yes Patrol Read Rate : Yes Alarm Control : Yes Cluster Support : No BBU : No ============> indicates BBU is not supported. However the BBU is present in this system for Castor card. Also MegaCli indicates that BBU is detected in the sytsem: MegaCli HW Config shows correctly that BBU is present. HW Configuration ================ SAS Address : 500605b0023142f0 BBU : Present ============> BBU present Alarm : Present NVRAM : Present Serial Debugger : Present Memory : Present Flash : Present Memory Size : 512MB TPM : Absent On board Expander: Absent Upgrade Key : Absent
Resolution:	Controller BBU capability is determine by "isCtrlBatteryPresent" defined in NVDATA instead of battery present.
Customer Defect Track No:	CSCti85988
Fix Impact:	Low
External Behavior Change:	No
Suggested Testing:	Check BBU capability with 2108 controllers (with or without battery present) Check BBU capability with 2008 controllers.
Child Tasks:	LSIP200060778

DFCT ID:	LSIP200167654
Headline:	(CL LSIP200159575) DC found when rebuild is progressing on R1 VD with media error under IO.
Description:	Miscompare is found. When R1 contain media errors and rebuilding on R1 under IO. I execute the TAS script. For more info please see the attached file. System Configuration is: --------------------------------------- Server: Dell PowerEdge 2900 Processor: (Intel(R) Xeon(R) CPU E5420 @ 2.50GHz) System RAM: 8GB System BIOS: 2.2.6 Operating System: Window 2K3_x86 SP2 Driver: 4.35.0 --------------------------------------- Controller: MegaRAID SAS MR9280-8e Pakage version: 12.13.0-0038 APP: 2.130.03-1222 MSM: 8.30-02 TAS: 00.03.12.00 --------------------------------------- Enclosure: Camden ESM Firmware: 0319 Drives: SAS
Steps to Reproduce:	1) Cascading 2 enclosures to port 1. 2) Create 12 R1 virtual drive with 2PDs in each group with default property. Each R1 across both enclosures - one drive per enclosure. 3) Start I/O with IO-monkey. 3) Start Error Injection script to keep injecting errors 3 different drives from 3 different R1s on first enclosure. 4) Then start power on/off script to cycle the last enclosure.
Resolution:	wrong cache line was getting invalidated
Suggested Testing:	Please see repro steps
Child Tasks:	LSIP200060816

DFCT ID:	LSIP200153500
Headline:	9280-4i4e unable to import a Foreign Configuration
Description:	Whenan array is created on a 9280-4i4e with SED drives, and then moved (External JBOD) to another 9280-4i4e, even with AutoEnhancedImport set to enable the Array will not import. Even when the array is put back on the original controller, it cannot be imported. I need to understand what this error is and how to get around it. Customer needs to have this configuration finalized by March 1st. Here are the FWTermlog errors (full FWTermlog is attached as well): EnclId - 13 SlotId - 0 Cannot import physical drive: Configuration conflict. The drive's place may have been taken by another drive, it could be part of another configuration, or the drive's configuration may be incompatible. Import or clear any other configurations that exist and then re-scan or return the drive to the original controller. EnclId - 13 SlotId - 1 Cannot import physical drive: Configuration conflict. The drive's place may have been taken by another drive, it could be part of another configuration, or the drive's configuration may be incompatible. Import or clear any other configurations that exist and then re-scan or return the drive to the original controller. EnclId - 13 SlotId - 2 Cannot import physical drive: Configuration conflict. The drive's place may have been taken by another drive, it could be part of another configuration, or the drive's configuration may be incompatible. Import or clear any other configurations that exist and then re-scan or return the drive to the original controller.
Steps to Reproduce:	Create an Array on one 9280-4i4e with SED drives. Move array to another 9280-4i4e and try to import.
Resolution:	This issue happens when a system is shut down, and part of the configuration is migrated to another system and followed by bring that configuration back. When the original system is finally booted, it will recognize that these drives but those drives are no longer native. This is because when these drives are imported elsewhere, their CTLR_GUID is changed.Also because of this, we don't end merging the DDF of PDs for these drives. So PD's DDF is uninitialized, hence the PD_GUID match from DDF PD records fails. But associated PD ref with other PD does succeed, as these drives are part of the DDF PD records in the COD on the reference Pd. As a fix, these drives need to be marked foreign here.
Suggested Testing:	1. create single drive R0 and 3 drive R5 on system A. 2. shutdown system A. 3. Create single drive R0 on system B. 4. shutdown system B. 5. Bring 3 drive R5 from system A to system B. 6. Start system B and import R5 to system B. 7. shutdown system B. 8. Bring that 3 drive R5 to system A. 9. start system A. The R5 could not be imported before the fix. Now the R5 can be imported.
Child Tasks:	LSIP200060776

Task ID:	LSIP200060778
Headline:	60778 MegaCli -AdpAllInfo shows BBU is not supported for Castor cards
Description:	MegaCli -AdpAllInfo shows BBU is not supported for Castor cards. This is a Castor with a BBU on it in an Alpine system.
State:	Delivered
Change Set Files:
References:	LSIP200156699(DFCT)

Task ID:	LSIP200060781
Headline:	60781 Label of Virtual Disk not provided by Drive Identify Command(LSIP200166039)
Description:	MegaRAID firmware does not support a mechanism of sending/storing the user defined label of a Virtual Drive Customer uses 3 RAID ctrls in one system. On each RAID ctrl the customer uses multiple Disk groups and multiple virtual drives on each disk group. To differentiate between the virtual drives the LINUX command lsscsi provides following output for a competitor product. lsscsi : [0:0:1:0] disk Adaptec VD_n V1.0 /dev/sda VD_n is the label of the generated virtual drive which has been entered during the creation of the virtual drive. Using the same command with MegaRAID and latest released SW package we do not output the name of the virtual drive, we generate a output using the default name of the MegaRAID controller: [0:2:0:0] disk LSI MR9280-24i4e 2.11 /dev/sda With MegaRAID the customer has no way to differentiate between the virtual drives. The customer plans to use 3 highport controllers 9280-24i4e / system and multiple virtual drives / controller. Handling >10 virtual drives without the corresponding label of the virtual drive is an issue for their manufacturing line. The customer plans to use 3 highport controllers 9280-24i4e / system and multiple volumes / controller. Entering lsscsi should provide the Label of each volume which requires the firmware to support a mechanism of sending/storing the user defined label of a RAID volume. Examples: IOMeter does not display the Label Names, for 10 different virtual drives on different disk groups (RAID5, RAID0 etc) we display 10 times exactly the same virtual drive label ! Setting up VDs in Filesystems: If you want to format a Virtual Disk, we do not provide any kind of info on the virtual disk, just the capacity is visible to the user. 10M09 is the latest released package not supporting this request.
State:	Delivered
Change Set Files:
References:	LSIP200166039(DFCT)

Task ID:	LSIP200060794
Headline:	60794 Log corruption due to Flash layout change
Description:	Fix the isse of FlashLogSize (-TTYLogSize) Since FW added the "ttyLogInflash" feature, we have to change the event log size only for MR. Also,the event log size for iMR was 2MB which is wrong. The correct size is 1MB.
State:	Delivered
Change Set Files:
References:	LSIP200157150(DFCT)

Task ID:	LSIP200060484
Headline:	60484 Red LED is lit on the next empty slot in an AIC/Xtor JBOD (LSIP200125226)
Description:	Bug reported on 8780 & 9280 only fixed on 8780, still issue with Post 10M09 GCA Patch Release 4.6 DFCT ID: LSID100126313 Headline: Red LED is lit on the next empty slot in an AIC/Xtor JBOD if the empy slot is immediatly after an array member Description: I have reproduced this on the 8880 with FW 1.40.42-0615 as well as the 9280-8E with FW 2.0.03-0635 under FC6 32bit Linux, 4.09 driver. This is NOT reproducible on the 8880 with FW 1.40.02-0514. The JBOD is an AIC/Xtor 3U 16 bay with the latest FW 1.14H.08, drives used were HDS SATA 3GB. I have recreated this numerous ways. Simplest steps to reproduce; 1. Insert three SATA disks in the array, slots 1-3. 2. MegaCli cfgclr a0 3. MegaCli cfgallfreedrv r5 a0 4. RED LED on slot 4 will be on. Tested R0/5 3,4,7 & 12 drives. Can also create a 1 drv R0 on slot 12 and slot 13 light Red will go on. Creating a 1 drv R0 on a slot with a drive present in the next slot does NOT reproduce the issue. Debug output is attached but I did not see any indication of an issue. This was originally reported to me by an AIC Eng. Manager whom I was working on a separate issue, his contact issue is below. He claims to have multiple customers reporting this issue and claims some are in stop ship because of the false error condition that it presents. This was originally reported with SAS & SATA drives on the 8888. Version of Bug Reported: 1.40.42-0615 / 2.0.03-0635 Version of Bug Fixed: 1.40.72-0702_PL-Ver-1.25.77.0 Steps to Reproduce: I have recreated this numerous ways. Simplest steps to reproduce; 1. Insert three SATA disks in the array, slots 1-3. 2. MegaCli cfgclr a0 3. MegaCli cfgallfreedrv r5 a0 4. RED LED on slot 4 will be on. Tested R0/5 3,4,7 & 12 drives. Can also create a 1 drv R0 on slot 12 and slot 13 light Red will go on. Creating a 1 drv R0 on a slot with a drive present in the next slot does NOT reproduce the issue. Resolution: Fixed Resolution Description: The bayNumber was retrieved wrongly when read from the additional element status descriptor page for the case of EIP bit is set to ‘1‚. The bayNumber is calculated correctly (at line 599) of file mapSes.c and the total disk device counter was used to do the calculation. Customer List: LSI -- LSI Fix Impact: Medium Suggested Testing: Please test with AIC/XTOR Jbod encl Child Tasks: LSID100152880
State:	Delivered
Change Set Files:
References:	LSIP200125226(DFCT)

Task ID:	LSIP200060796
Headline:	60796 Fw log should indicate if a created or existing VD is already initialized
Description:	adding an additional variable in the print statement in “LdDebug()”
State:	Delivered
Change Set Files:
References:	LSIP200166899(DFCT)

Task ID:	LSIP200060487
Headline:	60487 enable standby immediate command for HDDs for particular customer
Description:	enable standby immediate command for HDDs for particular customer
State:	Delivered
Change Set Files:
References:	LSIP200167677(CO)

Task ID:	LSIP200060836
Headline:	60836 Release Task 03/25/2011
Description:	Release Task
State:	Active
Change Set Files:
References:

Task ID:	LSIP200060486
Headline:	60486 enable device discovery prints for customer test fw builds
Description:	customer wants discovery prints in fw log automatically enabled for their test fw builds
State:	Delivered
Change Set Files:
References:	LSIP200164222(CO)

Task ID:	LSIP200060818
Headline:	60818 Tag uncorrectable ECC errors that occur after correctable ones appropriately
Description:	Function initFixEccErrors() checks for CHECK_MEMORY_ECC_NO_ERRORS but does not differentiate between correctable and uncorrectable errors. If errors were correctable in the first pass but not in the second pass of initCheckMemoryForEccErrors(); it will not be tagged as uncorrectable error. Cache will not be discarded and appropriate event will not be generated in that case. Also minor changes to be made in print statements for clarity
State:	Delivered
Change Set Files:
References:	LSIP200153454(DFCT)

Task ID:	LSIP200060816
Headline:	60816 Data corruption found when rebuild is progressing on R1 VD with media error under IO.
Description:	During the rebuild of R 1 volume, cache line would always be allocated for DATA ARM. Logic in BBM_PunctureBadBlocks would compute line index as 1 for a puncture entry for mirror arm. This would result in invalidating a wrong cache line. this is a back port from CC2.0
State:	Delivered
Change Set Files:
References:	LSIP200167654(DFCT)

Task ID:	LSIP200060795
Headline:	60795 Enable/disable of HDD cache doesn't appear to work( LSIP200164210 )
Description:	Code error in pd cache setting path does not allow f/w to set the pd cache setting.
State:	Delivered
Change Set Files:
References:	LSIP200164210(DFCT)

Task ID:	LSIP200060776
Headline:	60776 9280-4i4e unable to import a Foreign Configuration(LSIP200153500)- Enable import of foreign configuration during back t
Description:	When set of RAID drives are roamed back to back between 2 systems, import may fail because DDF will not be merged due to drive match with the native configuration. Mark these drives that are present in native configuration as foreign appropriately.
State:	Delivered
Change Set Files:
References:	LSIP200153500(DFCT)

Task ID:	LSIP200060711
Headline:	60711 Release Task 03/15/2011
Description:	Release Task
State:	Active
Change Set Files:
References:

Task ID:	LSIP200060792
Headline:	60792 SMART predictive failure detection doesn't happen in 10M09, for SATA drives
Description:	In 10M09, the additional feature to monitor drive temperature along with Predictive failure information has been added. As per this change, during every SMART polling cycle, LOG_SENSE(0x4D) command is issued to all disks with LOG_SENSE_INFORMATIONAL_EXCEPTIONS_PAGE page code(0x2F).PL after issuing the appropriate commands down to the drives, gives back SMART ASC/ASCQ and most recent temperature Information. For drives, where this page is not supported, this command fails with ILLEGAL_REQUEST, and the HDD temp is set to 0xFF(PD_INVALID) in PdInfo. From then onwards, FW shall revert back to using Request Sense for SMART Predictive failure indication. For SAS drives, this implementation seems to be working fine and FW switches to “REQUEST SENSE” once the “LOG_SENSE” command fails with SN_ILLEGAL_REQUEST & INVALID_FIELD_IN_CDB(5/24/00) . PL and Alta are expected to use SATL translation and support this LOG_SENSE page, for SATA drives. For SATA drives, the current implementation is using STP command ATA_CMD_READ_LOG_EXT, and getting the temperature information from the drive but misses getting the predictive Failure information. Fix: Temp monitoring functionality has been isolated from the Smart Polling Logic. Now, smartPollissue fires REQUEST_SENSE to the drive and in the callbackfunction, the new function smartPdTempMonitor() is invoked. This new function issues, ATA_CMD_READ_LOG_EXT( STP command) to the SATA drives for getting the temperature information, and issues LOG-SENSE command for SAS drives . If these commands fail to the drives, the drive temperature is set to 0xFF(PD_INVALID) in PdInfo and from that cycle onwards, this command will never be issued to the drive again.
State:	Delivered
Change Set Files:
References:	LSIP200146481(DFCT)     LSIP200146484(DFCT)

Task ID:	LSIP200060483
Title:	rebase bibhud_FW_SAS_10M09_Dev_Main on 20110214.110449.

Task ID:	LSIP200060469
Title:	rebase lstepan_FW_SAS_10M09_Dev_Main on 20110210.104603.

Task ID:	LSIP200060811
Title:	deliver atsushiy_FW_SAS_10M09_Dev_Main on 20110324.152033.

Task ID:	LSIP200060809
Title:	rebase bibhud_FW_SAS_10M09_Dev_Main on 20110324.180818.

Task ID:	LSIP200060784
Title:	rebase lstepan_FW_SAS_10M09_Dev_Main on 20110321.164043.

Task ID:	LSIP200060799
Title:	rebase rodneyp_FW_SAS_10M09_Dev_Main on 20110324.112735.

Task ID:	LSIP200060804
Title:	deliver rodneyp_FW_SAS_10M09_Dev_Main on 20110324.143152.

Task ID:	LSIP200060810
Title:	deliver bibhud_FW_SAS_10M09_Dev_Main on 20110324.181633.

Task ID:	LSIP200060829
Title:	deliver bibhud_FW_SAS_10M09_Dev_Main on 20110325.153846.

Task ID:	LSIP200060783
Title:	rebase snanduri_FW_SAS_10M09_Dev_Main on 20110321.160030.

Task ID:	LSIP200060835
Title:	deliver FW_SAS_10M09_Dev_Main on 20110325.171716.

Task ID:	LSIP200060797
Title:	deliver snanduri_FW_SAS_10M09_Dev_Main on 20110323.174920.

Task ID:	LSIP200060806
Title:	deliver lstepan_FW_SAS_10M09_Dev_Main on 20110324.124304.

Task ID:	LSIP200060435
Title:	rebase rodneyp_FW_SAS_10M09_Dev_Main on 20110207.153744.

Task ID:	LSIP200060834
Title:	deliver jayantd_FW_SAS_10M09_Dev_Main on 20110325.163927.

DFCT ID:	LSIP200157087
Headline:	EFI WebBIOS does not show product name/version for second adapter in adapter selection screen
Description:	EFI WebBIOS does not show product name/version for second adapter in adapter selection screen
Version of Bug Reported:	6.0-e_29-Rel
Steps to Reproduce:	Add multiple MR controllers with latest FW package to system Boot to EFI shell Boot to EFI WebBIOS by typing "drvcfg -s [DH#]" and then select option 1 Notice that the first adapter is displayed fine, but the second adapter is grayed out where the product name and version are displayed.
Resolution:	For displaying "Product Name" and "FW version" in EFI-WebBIOS, we are getting success only for one controller. Now we are returning success for all the attached controllers.
Fix Impact:	Low
External Behavior Change:	No
Suggested Testing:	Please refer the reproSteps.
Child Tasks:	LSIP200130463

DFCT ID:	LSIP200156291
Headline:	can't create DG with 8 PDs with enableReducedFeatureSet
Description:	you can't create a DG with 8 PDs. The issue was found with the WebBIOS 4.0-39-2_27-Rel. # of PD Create DG 7 Allowed 8 Disallowed
Steps to Reproduce:	1. Use iMR product, make sure "enableReductedFeatureSet = 1". 2. Boot system, go into WebBIOS. 3. Select 8 PDs and create DG. 4. WebBIOS won't let you create DG. you can do it up to 7 PDs, but not 8PDs.
Resolution:	check for allowed pds on the controller while adding into a DG.
Fix Impact:	Low
External Behavior Change:	No
Suggested Testing:	please refer the reprosteps.
Child Tasks:	LSIP200130160

DFCT ID:	LSIP200139977
Customer DFCT No:	7010593
Headline:	Option for populating the VD size field automatically
Description:	On the Virtual Drive Definition page in WebBIOS, the Select Size option is currently left for the user to fill in. In most cases the user must select the RAID Level and then look up the maximum size for that level under the Virtual Drives column on the right side. Possible user error and confusion could be minimized by adding a button to select the maximum size for desired level. So, after the user selects one of the possible levels for the chosen configuration, they could press this button to fill in the optimal size.
Steps to Reproduce:	Configure a VD in WebBios At VD details page, size of VD is blank
Resolution:	Provided "Update Size" button so that user can populate the corresponding RAID level size.
Customer Defect Track No:	7010593
Fix Impact:	Low
External Behavior Change:	No
Suggested Testing:	Please see the ReproSteps.
Child Tasks:	LSIP200128546

Task ID:	LSIP200090340
Headline:	90340 2108V_10M09_rel_version_6.0-26-e_21-Rel
Description:	2108V_10M09_rel_version_6.0-26-e_21-Rel
State:	Active
Change Set Files:
References:

Task ID:	LSIP200130463
Headline:	130463 LSIP200157087:->EFI WebBIOS does not show product name/version for second adapter in adapter selection screen
Description:	EFI WebBIOS does not show product name/version for second adapter in adapter selection screen
State:	Delivered
Change Set Files:
References:	LSIP200157087(DFCT)

Task ID:	LSIP200128546
Headline:	128546 Back porting of PR#LSIP200012152:Array Size During Creation from 11M02
Description:	Back porting of PR#LSIP200012152:Array Size During Creation from 11M02
State:	Delivered
Change Set Files:
References:	LSIP200139977(DFCT)

Task ID:	LSIP200130160
Headline:	130160 LSIP200156291:->can't create DG with 8 PDs with enableReducedFeatureSet
Description:	can't create DG with 8 PDs with enableReducedFeatureSet
State:	Delivered
Change Set Files:
References:	LSIP200156291(DFCT)

Task ID:	LSIP200160066
Title:	deliver avkumar_HWR_SAS_WEBBIOS_10M09 on 20110322.135741.

Task ID:	LSIP200128556
Title:	deliver avkumar_HWR_SAS_WEBBIOS_10M09 on 20110204.105946.

Task ID:	LSIP200130164
Title:	deliver avkumar_HWR_SAS_WEBBIOS_10M09 on 20110304.140509.

Task ID:	LSIP200109788
Title:	deliver avkumar_HWR_SAS_WEBBIOS_10M09 on 20101221.124114.

Task ID:	LSIP200128457
Title:	rebase avkumar_HWR_SAS_WEBBIOS_10M09 on 20110203.183138.

PR ID/CR ID:	LSIP200122625
Headline:	Adding New Customer IDs and XML for 2108 MROMB
Description:	1) One line description of the customer issue (customer pain point). Adding New Customer PnP IDs and XML files for 2108 MROMB - Customer: Quanta Computer Inc, IDs: (SVID): 152D (hex), 5421(Decimal) (SSID): 8985 (Hex) 2) List of affected components. Indicate pertinent HW, pre-boot apps, CLI, UX, etc. FW, Driver and other components that requires IDs. 3) Describe the current customer use cases. How is the product used today by this customer? What problems are encountered? Attach any pertinent logs and supporting information. N/A 4) Describe the desired customer use cases. What behavior does the customer wish to see? How, specifically, does it differ from current behavior? What components does the customer request to be changed? Will any desired customization be under customer control, or remain under LSI control? N/A 5) Does OEM require customization (e.g., customized behavior)? State the level of customization (i.e., does the level of control remain w/ OEM (e.g., NVDATA) and/or w/ END-USER (e.g., controller properties)?) Nope customization is required. 6) Urgency. Describe the relative severity of the problem, in the customer‚s view. Are there time-critical considerations to be aware of? Important. 7) Other supporting information (e.g., system vendor & model, OS and version, component revision levels, version of DAS component.) XML file attached.
State:		Implementation

IMP ID/CO ID:	LSIP200123293
Headline:	(auto4comp FW_SAS_LSI) Adding New Customer IDs and XML for 2108 MROMB
Description:	Auto Create from PR LSIP200122625. 1) One line description of the customer issue (customer pain point). Adding New Customer PnP IDs and XML files for 2108 MROMB - Customer: Quanta Computer Inc, IDs: (SVID): 152D (hex), 5421(Decimal) (SSID): 8985 (Hex) 2) List of affected components. Indicate pertinent HW, pre-boot apps, CLI, UX, etc. FW, Driver and other components that requires IDs. 3) Describe the current customer use cases. How is the product used today by this customer? What problems are encountered? Attach any pertinent logs and supporting information. N/A 4) Describe the desired customer use cases. What behavior does the customer wish to see? How, specifically, does it differ from current behavior? What components does the customer request to be changed? Will any desired customization be under customer control, or remain under LSI control? N/A 5) Does OEM require customization (e.g., customized behavior)? State the level of customization (i.e., does the level of control remain w/ OEM (e.g., NVDATA) and/or w/ END-USER (e.g., controller properties)?) Nope customization is required. 6) Urgency. Describe the relative severity of the problem, in the customer‚s view. Are there time-critical considerations to be aware of? Important. 7) Other supporting information (e.g., system vendor & model, OS and version, component revision levels, version of DAS component.) XML file attached.
State:	Ready_for_Delivery
Associated Task:	LSIP200160045
Task References:	LSIP200122625

DFCT ID:	LSIP200137750
Headline:	iMR: Nvdata change request for iMR to enable PowerSaving for Lynx2
Description:	Nvdata change request for iMR to enable PowerSaving for Lynx2
Steps to Reproduce:	n/a
Child Tasks:	LSIP200060725

Task ID:	LSIP200160045
Headline:	160045 add newxml file to nvdata
Description:	add newxml file to nvdata
State:	Delivered
Change Set Files:
References:	LSIP200123293(CO)

Task ID:	LSIP200160385
Headline:	160385 updating iMR nvdata image for 10M09 build
Description:	updating iMR nvdata image for 10M09 build for OEM Thor changes
State:	Active
Change Set Files:
References:

Task ID:	LSIP200060725
Headline:	60725 new 10M09 nvdata power savings setting for Lynx2
Description:	new 10M09 nvdata powerSave setting for Lynx2 changing the following for Lynx2... disablePowerSavings from FF to 00 PEReplyDelay from 18 to 00 delayPOST from 00 to 01
State:	Delivered
Change Set Files:
References:	LSIP200137750(DFCT)

Task ID:	LSIP200060782
Title:	rebase rodneyp_10M09_NVDATA_MR-iMR2_Dev_Main on 20110321.173242.

Task ID:	LSIP200160383
Title:	deliver 10M09_NVDATA_MR-iMR2_Dev_Main on 20110329.002807.

Task ID:	LSIP200160136
Title:	deliver rodneyp_10M09_NVDATA_MR-iMR2_Dev_Main on 20110322.210223.