Re: Adaptec AHD U320 operating as only U160 (take care with this hack!)

On 2/18/08, JoaoBR <joao_at_matik.com.br> wrote:
> On Monday 18 February 2008 05:01:01 JoaoBR wrote:
> > On Sunday 17 February 2008 14:23:41 Gelsema, P (Patrick) - FreeBSD wrote:
> > > On Sun, February 17, 2008 07:33, Justin T. Gibbs wrote:
> > > > Niki Denev wrote:
> > > >  > I was playing around with DTrace, tracing cam/xpt and the ahd driver
> > > >
> > > > and
> > > >
> > > >  > found out that if i comment the following code :
> > > >  >
> > > >  >         if ((spi3caps & SID_SPI_IUS) == 0)
> > > >  >                 spi->ppr_options &= ~MSG_EXT_PPR_IU_REQ;
> > > >  >
> > > >  > at line 6655 in sys/cam/cam_xpt.c my disks again negotiate as U320 :
> > > >  >
> > > >  >   da0 at ahd0 bus 0 target 0 lun 0
> > > >  >   da0: <SEAGATE ST336807LW 0C01> Fixed Direct Access SCSI-3 device
> > > >  >   da0: 320.000MB/s transfers (160.000MHz DT, offset 63, 16bit)
> > > >  >   da0: Command Queueing Enabled
> > > >  >   da0: 35003MB (71687372 512 byte sectors: 255H 63S/T 4462C)
> >
>
>
>
> take care with this hack because at least on Tyans and SM Mbs this hack does
> nothing good and cause continuous card dumps and some machines do not boot
> correctly (seems it depends on kind of disk installed), it is not usable
> here!
>
>
> the only card which stands this is the pci-x adaptec 29320
>
>
>
>
> > it is not working on Tyan MBs embedded u320 Adaptec, I still get with
> > complete world and kernel build on sources from this night
> >
> > #d3a 0L:a un<cShEeAdG!AT
> > E ST373207LW 0005> Fixed Direct Access SCSI-3 device
> > da0: 160.000MB/s transfers (80.000MHzS MDPT:,  AoPf fCsPeUt  #623 ,L
> > au1n6cbhietd)!
> >
> > da0: Command Queueing Enabled
> > da0: 70007MB (143374744 512 byte sectors: 255H 63S/T 8924C)
> > da1 at ahd0 bus 0 target 1 lun 0
> > da1: <SEAGATE ST373207LW 0003> Fixed Direct Access SCSI-3 device
> > da1: 160.000MB/s transfers (80.000MHz DT, offset 63, 16bit)
> > da1: Command Queueing Enabled
> > da1: 70007MB (143374744 512 byte sectors: 255H 63S/T 8924C)
> > Trying to mount root from ufs:/dev/da0s1a
> >
> >
> > It is this controller:
> >
> > ahd0: <Adaptec AIC7902 Ultra320 SCSI adapter> port
> > 0x9000-0x90ff,0x8800-0x88ff mem 0xfc9fc000-0xfc9fdfff irq 24 at device 6.0
> > on pci2
> > ahd0: [ITHREAD]
> > aic7902: Ultra320 Wide Channel A, SCSI Id=7, PCI-X 67-100Mhz, 512 SCBs
> > ahd1: <Adaptec AIC7902 Ultra320 SCSI adapter> port
> > 0x9800-0x98ff,0x9400-0x94ff mem 0xfc9fe000-0xfc9fffff irq 25 at device 6.1
> > on pci2
> > ahd1: [ITHREAD]
> > aic7902: Ultra320 Wide Channel B, SCSI Id=7, PCI-X 67-100Mhz, 512 SCBs
> >
> > which on first site seems to be the same on Supermicros MB:
> >
> > ahd0: <Adaptec AIC7902 Ultra320 SCSI adapter> port
> > 0xa800-0xa8ff,0xa400-0xa4ff mem 0xfc9fe000-0xfc9fffff irq 24 at device 3.0
> > on pci2
> > ahd0: [ITHREAD]
> > aic7902: Ultra320 Wide Channel A, SCSI Id=7, PCI-X 67-100Mhz, 512 SCBs
> > ahd1: <Adaptec AIC7902 Ultra320 SCSI adapter> port
> > 0xa000-0xa0ff,0xac00-0xacff mem 0xfc9fc000-0xfc9fdfff irq 25 at device 3.1
> > on pci2
> > ahd1: [ITHREAD]
> > aic7902: Ultra320 Wide Channel B, SCSI Id=7, PCI-X 67-100Mhz, 512 SCBs
> >
> >
> >
> >
> >
> > But I can confirm that it seems to be ok on 29320 Adaptec cards and on
> > Supermicro embedded U320 Adaptec
>
> --
>
> João
>

Yes, that's a gross hack, no doubt about it.
I posted it mainly to serve as a pointer to the
real problem, and I think it did that.
With version 1.30 of aic79xx_osm.c my negotiation
problems are gone.

  --Niki