Re: A question on ncr53c8xx driver

Robert (dnorman@cisco.com)
Sat, 13 Mar 1999 23:52:28 -0500 (EST)


I had this problem with the first LVD array I setup. I realized that I
didin't have a scsi-2 cable, but a normal SCSI UW cable. The kernel was
reporting the drives as 80MB/s drives, but I was getting them set to
20MB/s

--Drew

On Fri, 12 Mar 1999, Gerard Roudier wrote:

>
>
> On Thu, 11 Mar 1999, H.J. Lu wrote:
>
> > Hi,
> >
> > I have a question on the ncr53c8xx driver. I have one ncr53c895
> > and one ncr53c810a. I have 2 LVD drives on ncr53c895. For some
> > reason, they are set as FAST-20, not Fast-40 as I expected. The
> > BIOS sees them as 80MB/s. Is there the bug in the driver? Should
> > I try to take out ncr53c810a?
>
> >From driver point of view, the instance of the 895 has no relationship
> with the 810a one. A bug is always possible, but I doubt the problem would
> be due to a bug.
>
> LVD devices are in fact LVD/SE devices.
> LVD = low-voltage-differential (FAST40 a.k.a. Ultra2)
> SE = single-ended (FAST20 a.k.a. Ultra)
>
> All the devices on the SCSI BUS are sensing the DIFFSENS signal level and
> will switch to SE mode if this signal indicates that at least 1 device on
> the SCSI BUS is operating in SE mode, then limiting data transfer speed to
> FAST20.
>
> So, in order to use LVD signaling and then to get 80 MB/s sync wide data
> transfers on your SCSI BUS, you must ensure the following:
>
> - All devices including the controller are LVD capable and configured for
> this signaling (is such an option is possible).
> - All terminators (2) are LVD capable or LVD/SE (same behaviour as
> other devices).
> - All the cables meet the requirements for LVD.
>
> You also must:
> - Configure the driver for 40 MHz sync data transferts. Or tell it so at
> boot-up using 'ncr53c8xx=sync:10', if the driver has been configured for
> slower sync data transfers.
> - Configure your LVD devices for 80 MB/s data transfers in the SYMBIOS
> user set-up (NVRAM).
>
> It seems that LVD drive internal terminations are not LVD/SE but only SE.
> So, if any drive has its SE internal terminator enabled, then all the
> devices on the SCSI BUS will switch to SE mode, and the SCSI BUS will
> operate in FAST20 mode.
>
> Hope this will help.
>
> Regards,
> Gérard.
>
> PS: the driver should probably print out the SCSI mode which is currently
> operating on the BUS. I will add such a message in the next drivers'
> versions.
>
> > Thanks.
> >
> >
> > H.J.
> > ---
> > ncr53c8xx: at PCI bus 0, device 15, function 0
> > ncr53c8xx: 53c810a detected
> > ncr53c8xx: at PCI bus 0, device 16, function 0
> > ncr53c8xx: 53c895 detected with Symbios NVRAM
> > ncr53c895-0: rev=0x02, base=0xf4201400, io_port=0x1400, irq=19
> > ncr53c895-0: Symbios format NVRAM, ID 7, Fast-40, Parity Checking
> > ncr53c895-0: initial SCNTL3/DMODE/DCNTL/CTEST3/4/5 = (hex) 07/4e/a0/01/00/24
> > ncr53c895-0: final SCNTL3/DMODE/DCNTL/CTEST3/4/5 = (hex) 07/8e/a0/00/08/24
> > ncr53c895-0: on-chip RAM at 0xf4200000
> > ncr53c895-0: resetting, command processing suspended for 2 seconds
> > ncr53c895-0: restart (scsi reset).
> > ncr53c895-0: enabling clock multiplier
> > ncr53c895-0: Downloading SCSI SCRIPTS.
> > ncr53c810a-1: rev=0x12, base=0xf4201000, io_port=0x1000, irq=18
> > ncr53c810a-1: ID 7, Fast-10, Parity Checking
> > ncr53c810a-1: restart (scsi reset).
> > scsi0 : ncr53c8xx - revision 3.1h
> > scsi1 : ncr53c8xx - revision 3.1h
> > scsi : 2 hosts.
> > ncr53c895-0: command processing resumed
> > Vendor: QUANTUM Model: VIKING II 4.5WLS Rev: 5520
> > Type: Direct-Access ANSI SCSI revision: 02
> > Detected scsi disk sda at scsi0, channel 0, id 0, lun 0
> > Vendor: QUANTUM Model: QM318000TD-SW Rev: N491
> > Type: Direct-Access ANSI SCSI revision: 02
> > Detected scsi disk sdb at scsi0, channel 0, id 1, lun 0
> > ncr53c895-0-<0,0>: tagged command queue depth set to 8
> > ncr53c895-0-<1,0>: tagged command queue depth set to 8
> > Vendor: MICROP Model: 4110-09NB_Nov18F Rev: TN0F
> > Type: Direct-Access ANSI SCSI revision: 02
> > Detected scsi disk sdc at scsi1, channel 0, id 0, lun 0
> > Vendor: HP Model: C3725S Rev: 6039
> > Type: Direct-Access ANSI SCSI revision: 02
> > Detected scsi disk sdd at scsi1, channel 0, id 1, lun 0
> > ncr53c810a-1-<1,0>: tagged command queue depth set to 8
> > scsi : detected 4 SCSI disks total.
> > ncr53c895-0-<0,*>: WIDE SCSI (16 bit) enabled.
> > ncr53c895-0-<0,*>: FAST-20 WIDE SCSI 40.0 MB/s (50 ns, offset 31)
> > SCSI device sda: hdwr sector= 512 bytes. Sectors= 8910423 [4350 MB] [4.4 GB]
> > ncr53c895-0-<1,*>: WIDE SCSI (16 bit) enabled.
> > ncr53c895-0-<1,*>: FAST-20 WIDE SCSI 40.0 MB/s (50 ns, offset 31)
> > SCSI device sdb: hdwr sector= 512 bytes. Sectors= 35566500 [17366 MB] [17.4 GB]
> > ncr53c810a-1-<0,*>: FAST-10 SCSI 10.0 MB/s (100 ns, offset 8)
> > SCSI device sdc: hdwr sector= 512 bytes. Sectors= 2053880 [1002 MB] [1.0 GB]
> > ncr53c810a-1-<1,*>: FAST-10 SCSI 10.0 MB/s (100 ns, offset 8)
> > SCSI device sdd: hdwr sector= 512 bytes. Sectors= 4194058 [2047 MB] [2.0 GB]
> > Partition check:
> > sda: sda1 sda2 sda3 sda4 < sda5 sda6 sda7 sda8 >
> > sdb: sdb1 sdb2 sdb3 sdb4
> > sdc: sdc1
> > sdd:Dev 08:30 Sun disklabel: bad magic 0000
> > unknown partition table
> >
> >
>
>
> -
> To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
> the body of a message to majordomo@vger.rutgers.edu
> Please read the FAQ at http://www.tux.org/lkml/
>

-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.rutgers.edu
Please read the FAQ at http://www.tux.org/lkml/