Problems with ATA_CAM support in RELENG_8
I am running FreeBSD/amd64 RELENG_8 on a Dell Optiplex 745. The hard drive in the system is SATA and I have "Normal", not "Legacy" SATA support enabled in the BIOS. (BIOS is V2.6.4.) I am assuming this will enable native AHCI mode for the drive. I built a kernel with ATA_CAM support, but for some reason the SATA drive is probing at slow, UDMA2, speeds: ada0 at ata2 bus 0 scbus0 target 0 lun 0 ada0: ATA-7 SATA 2.x device ada0: 33.300MB/s transfers (UDMA2, PIO 8192bytes) ada0: 152587MB (31250 512 byte sectors: 16H 63S/T 16383C) When I run "camcontrol identify" on the drive, it states the drive is capable of UDMA6 speeds: backup# camcontrol identify ada0 pass0: ATA-7 SATA 2.x device pass0: 33.300MB/s transfers (UDMA2, PIO 8192bytes) protocol ATA/ATAPI-7 SATA 2.x device model ST3160812AS firmware revision 3.ADJ serial number 5LS8PPDD cylinders 16383 heads 16 sectors/track 63 sector size logical 512, physical 512, offset 0 LBA supported 268435455 sectors LBA48 supported 31250 sectors PIO supported PIO4 DMA supported WDMA2 UDMA6 Feature Support EnableValue Vendor read ahead yes yes write cacheyes yes flush cacheyes yes overlapno Tagged Command Queuing (TCQ) no no Native Command Queuing (NCQ) yes 32 tags SMART yes yes microcode download yes yes security yes no power management yes yes advanced power management no no 65278/0xFEFE automatic acoustic management yes yes 254/0xFE208/0xD0 media status notification no no power-up in Standbyno no write-read-verify no no 0/0x0 unload no no free-fall no no data set management (TRIM) no So, why the slower speed? Also, "camcontrol identify" states the drive supports NCQ with up to 32 tags supported, yet "camcontrol tags ada0" reports only 1 device opening, not 32 as I would expect: backup# camcontrol tags ada0 (pass0:ata2:0:0:0): device openings: 1 To enable ATA_CAM AHCI support, I included this in my kernel config file: # ATA and ATAPI devices options ATA_CAM device ahci device atacore device atapci options ATA_STATIC_ID # Static device numbering Is this the correct way to enable ATA_CAM AHCI support? I tried initially including just "options ATA_CAM" and "device ahci" but the resultant kernel would not probe my disk drive as ada0. Does my problem lie with my kernel config or is the Dell Optiplex 745 BIOS brain dead when it comes to AHCI native support? The only option it appears to have in the BIOS is "Normal" and "Legacy" when it comes to the SATA controller mode. I've attached my full kernel config and dmesg at the end of this e-mail. Cheers, Paul. Kernel config: # # GENERIC -- Generic kernel configuration file for FreeBSD/amd64 # # For more information on this file, please read the config(5) manual page, # and/or the handbook section on Kernel Configuration Files: # # http://www.FreeBSD.org/doc/en_US.ISO8859-1/books/handbook/kernelconfig-config.html # # The handbook is also available locally in /usr/share/doc/handbook # if you've installed the doc distribution, otherwise always see the # FreeBSD World Wide Web server (http://www.FreeBSD.org/) for the # latest information. # # An exhaustive list of options and more detailed explanations of the # device lines is also present in the ../../conf/NOTES and NOTES files. # If you are in doubt as to the purpose or necessity of a line, check first # in NOTES. # # $FreeBSD: src/sys/amd64/conf/GENERIC,v 1.531.2.13 2010/05/02 06:24:17 imp Exp $ cpu HAMMER ident VPN # To statically compile in device wiring instead of /boot/device.hints #hints "GENERIC.hints" # Default places to look for devices. # Use the following to compile in values accessible to the kernel # through getenv() (or kenv(1) in userland). The format of the file # is 'variable=value', see kenv(1) # # env "GENERIC.env" #makeoptionsDEBUG=-g# Build kernel with gdb(1) debug symbols options SCHED_ULE # ULE scheduler options PREEMPTION # Enable kernel thread preemption options INET# InterNETworking options INET6 # IPv6 communications protocols options SCTP# Stream Control Transmission Protocol options FFS # Berkeley Fast Filesystem options SOFTUPDATES # Enable FFS soft updates support options UFS_ACL # Support for
Re: Problems with ATA_CAM support in RELENG_8
On Wed, Jun 30, 2010 at 12:07:54PM -0400, Paul Mather wrote: > I am running FreeBSD/amd64 RELENG_8 on a Dell Optiplex 745. The hard drive > in the system is SATA and I have "Normal", not "Legacy" SATA support enabled > in the BIOS. (BIOS is V2.6.4.) I am assuming this will enable native AHCI > mode for the drive. That isn't correct. Some Dell OptiPlex systems explicitly have an AHCI option in their BIOS for enabling/disabling AHCI. I do not believe the 745 offers this capability, which means no AHCI is available. (We use these OptiPlex models at my workplace so I'm familiar with them) > I built a kernel with ATA_CAM support, but for some reason the SATA drive is > probing at slow, UDMA2, speeds: I don't know anything about this option, but the "common" way to get ATA to utilise SCSI CAM is to enable AHCI in one's system BIOS and load the ahci.ko (not ataahci.ko) driver using /boot/loader.conf (ahci_load="yes") or "load ahci" from the loader "OK" prompt. That's literally all you need. In your case since you don't have AHCI available, you're probably going to have to stick to using standard ata(4) (read: remove ATA_CAM and ahci from your kernel configuration). Be aware ahci.ko IS NOT the same thing as ataahci.ko. The latter adds AHCI support but via the classic ata(4) layer (meaning: AHCI without SCSI CAM. This also lacks NCQ support, but does gain you hot-swap capability). > Does my problem lie with my kernel config or is the Dell Optiplex 745 > BIOS brain dead when it comes to AHCI native support? The only option > it appears to have in the BIOS is "Normal" and "Legacy" when it comes > to the SATA controller mode. It's not "brain dead". "Normal" is not the same thing as AHCI. "Normal" tells the controller to run in standard SATA mode, "Legacy" causes the controller to emulate classic PATA (literally you'll see PATA controllers show up on IRQs 14 and 15, etc.). Validation that your system doesn't have AHCI enabled/doesn't have AHCI available is here: > atapci0: port > 0xfe00-0xfe07,0xfe10-0xfe13,0xfe20-0xfe27,0xfe30-0xfe33,0xfec0-0xfecf,0xecc0-0xeccf > irq 20 at device 31.2 on pci0 > atapci0: [ITHREAD] > ata2: on atapci0 > ata2: [ITHREAD] > ata3: on atapci0 > ata3: [ITHREAD] > atapci1: port > 0xfe40-0xfe47,0xfe50-0xfe53,0xfe60-0xfe67,0xfe70-0xfe73,0xfed0-0xfedf,0xecd0-0xecdf > irq 20 at device 31.5 on pci0 > atapci1: [ITHREAD] > ata4: on atapci1 > ata4: [ITHREAD] > ata5: on atapci1 > ata5: [ITHREAD] An AHCI-enabled controller, with a system using ahci.ko, will look like this: ahci0: port 0x1c50-0x1c57,0x1c44-0x1c47,0x1c48-0x1c4f,0x1c40-0x1c43,0x18e0-0x18ff mem 0xdc000800-0xdc000fff irq 17 at device 31.2 on pci0 ahci0: [ITHREAD] ahci0: AHCI v1.20 with 6 3Gbps ports, Port Multiplier supported ahcich0: at channel 0 on ahci0 ahcich0: [ITHREAD] ahcich1: at channel 1 on ahci0 ahcich1: [ITHREAD] ahcich2: at channel 2 on ahci0 ahcich2: [ITHREAD] ahcich3: at channel 3 on ahci0 ahcich3: [ITHREAD] ahcich4: at channel 4 on ahci0 ahcich4: [ITHREAD] ahcich5: at channel 5 on ahci0 ahcich5: [ITHREAD] -- | Jeremy Chadwick j...@parodius.com | | Parodius Networking http://www.parodius.com/ | | UNIX Systems Administrator Mountain View, CA, USA | | Making life hard for others since 1977. PGP: 4BD6C0CB | ___ freebsd-stable@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-stable To unsubscribe, send any mail to "freebsd-stable-unsubscr...@freebsd.org"
Re: Problems with ATA_CAM support in RELENG_8
On Jun 30, 2010, at 1:38 PM, Alexander Motin wrote: >> To enable ATA_CAM AHCI support, I included this in my kernel config file: >> >> # ATA and ATAPI devices >> options ATA_CAM >> device ahci >> device atacore >> device atapci >> options ATA_STATIC_ID # Static device numbering >> >> Is this the correct way to enable ATA_CAM AHCI support? I tried >> initially including just "options ATA_CAM" and "device ahci" but the >> resultant kernel would not probe my disk drive as ada0. > > Your controller seems to not report AHCI support. In such case legacy > mode driver attaches to it. But as soon as you have no `device ataintel` > line, only generic driver was there, limited by UDMA2 mode. > > PS: ATA_STATIC_ID is useless when ATA_CAM option enabled. Thank you (and Jeremy Chadwick) for the help and information. The kernel configuration options I used above were taken from a VirtualBox FreeBSD/amd64 install I have that I converted over to ATA_CAM when the code first went into RELENG_8 and it wasn't exactly clear at the time what options were absolutely required. (I'm not even sure that "options ATA_CAM" is needed any more, given "device ahci" implies it.) >> Does my problem lie with my kernel config or is the Dell Optiplex 745 >> BIOS brain dead when it comes to AHCI native support? The only option >> it appears to have in the BIOS is "Normal" and "Legacy" when it comes >> to the SATA controller mode. > > Your controller is not identified as AHCI: > >> atapci0: port >> 0xfe00-0xfe07,0xfe10-0xfe13,0xfe20-0xfe27,0xfe30-0xfe33,0xfec0-0xfecf,0xecc0-0xeccf >> irq 20 at device 31.2 on pci0 >> atapci0: [ITHREAD] >> ata2: on atapci0 >> ata2: [ITHREAD] >> ata3: on atapci0 >> ata3: [ITHREAD] >> atapci1: port >> 0xfe40-0xfe47,0xfe50-0xfe53,0xfe60-0xfe67,0xfe70-0xfe73,0xfed0-0xfedf,0xecd0-0xecdf >> irq 20 at device 31.5 on pci0 >> atapci1: [ITHREAD] >> ata4: on atapci1 >> ata4: [ITHREAD] >> ata5: on atapci1 >> ata5: [ITHREAD] > > You may check `pciconf -lvcb` output. For ICH8 with AHCI you should see > there something like: > ah...@pci0:0:31:2: class=0x010601 card=0xa00c14ff chip=0x28298086 > rev=0x03 hdr=0x00 >vendor = 'Intel Corporation' >device = 'Mobile SATA AHCI Controller' >class = mass storage >subclass = SATA >bar [10] = type I/O Port, range 32, base 0xe880, size 8, enabled >bar [14] = type I/O Port, range 32, base 0xe800, size 4, enabled >bar [18] = type I/O Port, range 32, base 0xe480, size 8, enabled >bar [1c] = type I/O Port, range 32, base 0xe400, size 4, enabled >bar [20] = type I/O Port, range 32, base 0xe080, size 32, enabled >bar [24] = type Memory, range 32, base 0xfeaff800, size 2048, enabled >cap 05[80] = MSI supports 4 messages enabled with 4 messages >cap 01[70] = powerspec 3 supports D0 D3 current D0 >cap 12[a8] = SATA Index-Data Pair > > Pay attention to "subclass = SATA". Then that's where my problem lies: atap...@pci0:0:31:2:class=0x01018f card=0x01da1028 chip=0x28208086 rev=0x02 hdr=0x00 vendor = 'Intel Corporation' device = 'SATA IDE Controller:4 port (82801HB/HR/HH/HO)' class = mass storage subclass = ATA bar [10] = type I/O Port, range 32, base 0xfe00, size 8, enabled bar [14] = type I/O Port, range 32, base 0xfe10, size 4, enabled bar [18] = type I/O Port, range 32, base 0xfe20, size 8, enabled bar [1c] = type I/O Port, range 32, base 0xfe30, size 4, enabled bar [20] = type I/O Port, range 32, base 0xfec0, size 16, enabled bar [24] = type I/O Port, range 32, base 0xecc0, size 16, enabled cap 01[70] = powerspec 3 supports D0 D3 current D0 atap...@pci0:0:31:5:class=0x010185 card=0x01da1028 chip=0x28258086 rev=0x02 hdr=0x00 vendor = 'Intel Corporation' device = '82801H (ICH8 Family) 2 port SATA Controller' class = mass storage subclass = ATA bar [10] = type I/O Port, range 32, base 0xfe40, size 8, enabled bar [14] = type I/O Port, range 32, base 0xfe50, size 4, enabled bar [18] = type I/O Port, range 32, base 0xfe60, size 8, enabled bar [1c] = type I/O Port, range 32, base 0xfe70, size 4, enabled bar [20] = type I/O Port, range 32, base 0xfed0, size 16, enabled bar [24] = type I/O Port, range 32, base 0xecd0, size 16, enabled cap 01[70] = powerspec 3 supports D0 D3 current D0 I thought ICH8 supported AHCI, but maybe it's only ICH8R that does? I'm assuming that "subclass = ATA" means the controller can't operate in AHCI mode. The BIOS setting is also confusing. It has two options, "Normal" and "Legacy." "Normal" mode says, "The hard drive controller is configured for native mode. This mode provides the highest drive performance and most flexibility." I guess I misinterpreted "native mode" to be AHCI mode. Thanks again for the help and for clearing things up. Cheer
Re: Problems with ATA_CAM support in RELENG_8
Paul Mather wrote: >> PS: ATA_STATIC_ID is useless when ATA_CAM option enabled. > > Thank you (and Jeremy Chadwick) for the help and information. The kernel > configuration options I used above were taken from a VirtualBox FreeBSD/amd64 > install I have that I converted over to ATA_CAM when the code first went into > RELENG_8 and it wasn't exactly clear at the time what options were absolutely > required. (I'm not even sure that "options ATA_CAM" is needed any more, > given "device ahci" implies it.) `options ATA_CAM` enables CAM wrapper for legacy drivers, which gave you adaX devices instead of adX. It doesn't give major benefits, just unifies behavior. > I thought ICH8 supported AHCI, but maybe it's only ICH8R that does? All ICH8 support AHCI, though not all BIOSes initialize them respectively. The only benefit of ICH8R is a port multipliers support, though it also seems to be artificial limitation - I've seen BIOS allowing it on regular ICH8. > I'm assuming that "subclass = ATA" means the controller can't operate in AHCI > mode. Yes. At least it is not configured to do so. > The BIOS setting is also confusing. It has two options, "Normal" and > "Legacy." "Normal" mode says, "The hard drive controller is configured for > native mode. This mode provides the highest drive performance and most > flexibility." I guess I misinterpreted "native mode" to be AHCI mode. "Native" mode means that ATA controllers works as normal PCI device. In "Legacy" mode controller uses legacy ISA I/O ports and IRQs. -- Alexander Motin ___ freebsd-stable@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-stable To unsubscribe, send any mail to "freebsd-stable-unsubscr...@freebsd.org"
Re: Problems with ATA_CAM support in RELENG_8
On Wed, Jun 30, 2010 at 04:52:22PM -0400, Paul Mather wrote: > I thought ICH8 supported AHCI, but maybe it's only ICH8R that does? No, not necessarily. There are ICHxx (non-ICHxxR) models which definitely provide AHCI (ex. ICHxxM for their Mobile chipsets; yes, many laptops do have AHCI capability). Thus there are many different sub-models of ICH8, in addition to different perf/chip revisions of ICH8. Specification details: http://www.intel.com/assets/pdf/datasheet/313056.pdf http://www.intel.com/design/chipsets/specupdt/313057.htm Dell very likely chose not to engineer this particular motherboard to not provide AHCI capability (Option ROM space (usually included with the system BIOS), PCI and ICH8 register setups (done by the BIOS), etc.)). I know Gigabyte has done this with some of their boards as well. If you have concerns that your motherboard provides features that you can't use, you need to contact Dell. > I'm assuming that "subclass = ATA" means the controller can't operate > in AHCI mode. The BIOS setting is also confusing. It has two > options, "Normal" and "Legacy." "Normal" mode says, "The hard drive > controller is configured for native mode. This mode provides the > highest drive performance and most flexibility." I guess I > misinterpreted "native mode" to be AHCI mode. "Native", "Normal", and "Enhanced" are all terms used to describe the controller operating in standard SATA (non-AHCI) fashion. Dell's in-BIOS description is nothing but schmooze; it's not the same as AHCI. "Legacy" just means the controller appears and operates purely in PATA-esque mode, so that you can use SATA drives on systems which don't have SATA chipset drivers (ex. MS-DOS). To get an idea of what an AHCI-capable board offers via the BIOS, you should review the following manual for a Supermicro ICH7-based system. This chipset is older than yours, chosen for sake of example. See Chapter 7-3, section "Main Setup Features": http://www.supermicro.com/manuals/superserver/1U/MNL-0813.pdf Furthermore, I can assure you that the Dell OptiPlex 745 Mini-Tower offers AHCI capability, and toggling its use via the system BIOS. You won't find it mentioned in Dell's "Tech Specs" PDF about this though. I realise this is not the system you have, but I'm simply pointing out that it's a board manufacturer decision whether or not to provide the capability, and that their decision obviously changed over the years. -- | Jeremy Chadwick j...@parodius.com | | Parodius Networking http://www.parodius.com/ | | UNIX Systems Administrator Mountain View, CA, USA | | Making life hard for others since 1977. PGP: 4BD6C0CB | ___ freebsd-stable@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-stable To unsubscribe, send any mail to "freebsd-stable-unsubscr...@freebsd.org"
Re: Problems with ATA_CAM support in RELENG_8
xc ___ freebsd-stable@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-stable To unsubscribe, send any mail to "freebsd-stable-unsubscr...@freebsd.org"
Re: Problems with ATA_CAM support in RELENG_8
On Jun 30, 2010, at 5:18 PM, Alexander Motin wrote: > Paul Mather wrote: >>> PS: ATA_STATIC_ID is useless when ATA_CAM option enabled. >> >> Thank you (and Jeremy Chadwick) for the help and information. The kernel >> configuration options I used above were taken from a VirtualBox >> FreeBSD/amd64 install I have that I converted over to ATA_CAM when the code >> first went into RELENG_8 and it wasn't exactly clear at the time what >> options were absolutely required. (I'm not even sure that "options ATA_CAM" >> is needed any more, given "device ahci" implies it.) > > `options ATA_CAM` enables CAM wrapper for legacy drivers, which gave you > adaX devices instead of adX. It doesn't give major benefits, just > unifies behavior. So, does that mean if you omit "options ATA_CAM" and have "device ahci" you will get adX devices, not adaX devices? In other words, if you have "device ahci" (or "device siis" or "device mvs") will you will always get adaX devices, whether or not you have "options ATA_CAM" in your kernel config file? Does "options ATA_CAM" work with "device ata" or the modular ATA subsystem? Is that the intended use of "options ATA_CAM": to provide adaX devices and a CAM interface for accessing ATA devices? Cheers, Paul.___ freebsd-stable@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-stable To unsubscribe, send any mail to "freebsd-stable-unsubscr...@freebsd.org"
Re: Problems with ATA_CAM support in RELENG_8
On Thu, Jul 01, 2010 at 08:26:31AM -0400, Paul Mather wrote:
> On Jun 30, 2010, at 5:21 PM, Jeremy Chadwick wrote:
> >> I'm assuming that "subclass = ATA" means the controller can't operate
> >> in AHCI mode. The BIOS setting is also confusing. It has two
> >> options, "Normal" and "Legacy." "Normal" mode says, "The hard drive
> >> controller is configured for native mode. This mode provides the
> >> highest drive performance and most flexibility." I guess I
> >> misinterpreted "native mode" to be AHCI mode.
> >
> > "Native", "Normal", and "Enhanced" are all terms used to describe the
> > controller operating in standard SATA (non-AHCI) fashion. Dell's
> > in-BIOS description is nothing but schmooze; it's not the same as AHCI.
> > "Legacy" just means the controller appears and operates purely in
> > PATA-esque mode, so that you can use SATA drives on systems which don't
> > have SATA chipset drivers (ex. MS-DOS).
>
> Thank you for helping clear that up. I have a Supermicro
> X8ST3-F-based server here, and it offers three choices for drive mode:
> IDE, RAID, and AHCI. I think it's much clearer, there, what you're
> getting. Dell could learn a thing or two from them. :-)
It's a matter of demographic -- Dell often caters to home/residential
users, while Supermicro caters to engineers, SAs, and other highly
technical folk. The latter excludes "gamers" or "tech forum freaks";
these two rarely, if ever, actually program or interface with the
devices or chips they're talking about. Speculation on my part, but I'm
sure if you Google for "ICH8 ACHI" you'll find hundreds of posts on
random forums, and most of these posts will do nothing but confuse. :-)
Dell started adding more technically-savvy options to their later
models, probably because of pressure from the above two groups, or
competitive pressure. Hard to say.
> > Furthermore, I can assure you that the Dell OptiPlex 745 Mini-Tower
> > offers AHCI capability, and toggling its use via the system BIOS. You
> > won't find it mentioned in Dell's "Tech Specs" PDF about this though. I
> > realise this is not the system you have, but I'm simply pointing out
> > that it's a board manufacturer decision whether or not to provide the
> > capability, and that their decision obviously changed over the years.
>
> Oddly enough, it *is* a Dell OptiPlex 745 in Mini-Tower configuration.
> This is the manual I'm led to via the service tag:
> http://support.dell.com/support/edocs/systems/op745/en/UG_en/mtindex.htm
>
> Are you saying there is an option in the BIOS to enable AHCI on these
> systems?
Sorry for inducing confusion -- I meant the Dell OptiPlex 760.
At my workplace we literally have 4 different Dell OptiPlex models in
use as workstations. I often mix up the model numbers, case in point.
I'm re-CC'ing the mailing list here, because I want to make sure my
mistake gets noted in case someone ends up searching the lists ("Hey! I
have a 745, what is this guy talking about?!?")
--
| Jeremy Chadwick j...@parodius.com |
| Parodius Networking http://www.parodius.com/ |
| UNIX Systems Administrator Mountain View, CA, USA |
| Making life hard for others since 1977. PGP: 4BD6C0CB |
___
freebsd-stable@freebsd.org mailing list
http://lists.freebsd.org/mailman/listinfo/freebsd-stable
To unsubscribe, send any mail to "freebsd-stable-unsubscr...@freebsd.org"
Re: Problems with ATA_CAM support in RELENG_8
Paul Mather wrote: > On Jun 30, 2010, at 5:18 PM, Alexander Motin wrote: >> Paul Mather wrote: PS: ATA_STATIC_ID is useless when ATA_CAM option enabled. >>> Thank you (and Jeremy Chadwick) for the help and information. The kernel >>> configuration options I used above were taken from a VirtualBox >>> FreeBSD/amd64 install I have that I converted over to ATA_CAM when the code >>> first went into RELENG_8 and it wasn't exactly clear at the time what >>> options were absolutely required. (I'm not even sure that "options >>> ATA_CAM" is needed any more, given "device ahci" implies it.) >> `options ATA_CAM` enables CAM wrapper for legacy drivers, which gave you >> adaX devices instead of adX. It doesn't give major benefits, just >> unifies behavior. > > So, does that mean if you omit "options ATA_CAM" and have "device ahci" you > will get adX devices, not adaX devices? In other words, if you have "device > ahci" (or "device siis" or "device mvs") will you will always get adaX > devices, whether or not you have "options ATA_CAM" in your kernel config file? If you have ahci, siis and mvs drivers in kernel, but no "options ATA_CAM", you will get adaX devices for native SATA controllers, supported by these three drivers and adX for other (legacy) ones. > Does "options ATA_CAM" work with "device ata" or the modular ATA subsystem? Both. It is an option of "device atacore", which is part of "device ata" > Is that the intended use of "options ATA_CAM": to provide adaX devices and a > CAM interface for accessing ATA devices? Yes. It disables ATA peripheral drivers and bus management code, allowing CAM ones to be used on top of the ATA controller drivers. -- Alexander Motin ___ freebsd-stable@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-stable To unsubscribe, send any mail to "freebsd-stable-unsubscr...@freebsd.org"
Re: Fwd: Problems with ATA_CAM support in RELENG_8
Garrett Cooper wrote: > I am running FreeBSD/amd64 RELENG_8 on a Dell Optiplex 745. The hard > drive in the system is SATA and I have "Normal", not "Legacy" SATA > support enabled in the BIOS. (BIOS is V2.6.4.) I am assuming this > will enable native AHCI mode for the drive. > > I built a kernel with ATA_CAM support, but for some reason the SATA > drive is probing at slow, UDMA2, speeds: > > ada0 at ata2 bus 0 scbus0 target 0 lun 0 > ada0: ATA-7 SATA 2.x device > ada0: 33.300MB/s transfers (UDMA2, PIO 8192bytes) > ada0: 152587MB (31250 512 byte sectors: 16H 63S/T 16383C) > When I run "camcontrol identify" on the drive, it states the drive is > capable of UDMA6 speeds: > > backup# camcontrol identify ada0 > PIO supported PIO4 > DMA supported WDMA2 UDMA6 > Native Command Queuing (NCQ) yes 32 tags > > So, why the slower speed? Also, "camcontrol identify" states the > drive supports NCQ with up to 32 tags supported, yet "camcontrol tags > ada0" reports only 1 device opening, not 32 as I would expect: > > backup# camcontrol tags ada0 > (pass0:ata2:0:0:0): device openings: 1 Speed, number of tags and other options limited by lowest of device and controller capabilities. > To enable ATA_CAM AHCI support, I included this in my kernel config file: > > # ATA and ATAPI devices > options ATA_CAM > device ahci > device atacore > device atapci > options ATA_STATIC_ID # Static device numbering > > Is this the correct way to enable ATA_CAM AHCI support? I tried > initially including just "options ATA_CAM" and "device ahci" but the > resultant kernel would not probe my disk drive as ada0. Your controller seems to not report AHCI support. In such case legacy mode driver attaches to it. But as soon as you have no `device ataintel` line, only generic driver was there, limited by UDMA2 mode. PS: ATA_STATIC_ID is useless when ATA_CAM option enabled. > Does my problem lie with my kernel config or is the Dell Optiplex 745 > BIOS brain dead when it comes to AHCI native support? The only option > it appears to have in the BIOS is "Normal" and "Legacy" when it comes > to the SATA controller mode. Your controller is not identified as AHCI: > atapci0: port > 0xfe00-0xfe07,0xfe10-0xfe13,0xfe20-0xfe27,0xfe30-0xfe33,0xfec0-0xfecf,0xecc0-0xeccf > irq 20 at device 31.2 on pci0 > atapci0: [ITHREAD] > ata2: on atapci0 > ata2: [ITHREAD] > ata3: on atapci0 > ata3: [ITHREAD] > atapci1: port > 0xfe40-0xfe47,0xfe50-0xfe53,0xfe60-0xfe67,0xfe70-0xfe73,0xfed0-0xfedf,0xecd0-0xecdf > irq 20 at device 31.5 on pci0 > atapci1: [ITHREAD] > ata4: on atapci1 > ata4: [ITHREAD] > ata5: on atapci1 > ata5: [ITHREAD] You may check `pciconf -lvcb` output. For ICH8 with AHCI you should see there something like: ah...@pci0:0:31:2: class=0x010601 card=0xa00c14ff chip=0x28298086 rev=0x03 hdr=0x00 vendor = 'Intel Corporation' device = 'Mobile SATA AHCI Controller' class = mass storage subclass = SATA bar [10] = type I/O Port, range 32, base 0xe880, size 8, enabled bar [14] = type I/O Port, range 32, base 0xe800, size 4, enabled bar [18] = type I/O Port, range 32, base 0xe480, size 8, enabled bar [1c] = type I/O Port, range 32, base 0xe400, size 4, enabled bar [20] = type I/O Port, range 32, base 0xe080, size 32, enabled bar [24] = type Memory, range 32, base 0xfeaff800, size 2048, enabled cap 05[80] = MSI supports 4 messages enabled with 4 messages cap 01[70] = powerspec 3 supports D0 D3 current D0 cap 12[a8] = SATA Index-Data Pair Pay attention to "subclass = SATA". -- Alexander Motin ___ freebsd-stable@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-stable To unsubscribe, send any mail to "freebsd-stable-unsubscr...@freebsd.org"
