Date: Thu, 25 Nov 1999 21:40:01 +0100 (CET) From: Wilko Bulte <wilko@yedi.iaf.nl> To: groudier@club-internet.fr (Gerard Roudier) Cc: cdillon@wolves.k12.mo.us, roberto@keltia.freenix.fr, scsi@FreeBSD.ORG Subject: Re: NEWS: SYM53C1010 and Ultra3 support Message-ID: <199911252040.VAA02381@yedi.iaf.nl> In-Reply-To: <Pine.LNX.3.95.991125213854.368A-100000@localhost> from Gerard Roudier at "Nov 25, 1999 9:50:37 pm"
next in thread | previous in thread | raw e-mail | index | archive | help
As Gerard Roudier wrote ... > On Wed, 24 Nov 1999, Wilko Bulte wrote: > > > As Gerard Roudier wrote ... > > > > > > > > > On Tue, 23 Nov 1999, Chris Dillon wrote: > > > > > > > On Wed, 24 Nov 1999, Ollivier Robert wrote: > > > > > > > > > According to Gerard Roudier: > > > > [...] > > > The SYM driver needs LOAD/STORE SCRIPTS instructions to be supported by > > > the PCI/SCSI chip. > > > As a result, 810 rev < 0x10, 825 rev < 0x10 and 815 all revisions support > > > is dropped. LOAD/STORE have lots of advantages (I have detailed some in > > > > Allow me to make a remark here: where does that leave people with older > > *mainboards with embedded NCR chips* ? Might be not too interesting > > for Intel folks, but most likely *is* for the Alpha folks. > > > > Mind you, I have not checked this on my Alphas (and the one Intel > > with embedded ncr) but I have a *bad* feeling over it as they are > > older stuff. > > > > The 'lets throw away the old hardware' that reigns the Intel world is > > less applicable to the Alpha world. A driver that cannot accomodate > > older hardware is a step in the wrong direction here. > > > > Just my $ 0.02 > > You just lost your money. There is no problem using the both drivers on I can afford $0.02 ;-) I was just worried about the maintenance on the ncr driver as it is only used on old hardware. But read on.. > The right direction is, in my opinion, to have drivers that allows users > to get IN TIME functionnalities and performances they pay for. By the way, > this not only requires development but also maintainance. > > I have decided to have 2 drivers for the SYM53C8XX family and, as you > should have known, nobody else seemed to have decided to do differently > for _actually_ providing users features users pay for and not having to > deal for a long time with to complex driver code. By the way, SYMBIOS did > the same, independantly. I'm not sure I understand completely what you are saying. I agree that one wants to have drivers that take full advantage of the features the hardware has to offer. So a nice new clean driver is a good thing. But from pulling the sym tarball over onto one of my Alpha's and reading the README (I know, should have done that earlier, but I was too busy with other stuff.. :-( ) I understand that you make the 'ncr' driver take the older chips and have the sym take the new ones. Being curious I put the sym into -current on the Alpha box and got: EB64+ Alpha 21064A Evaluation Board 274 MHz, 274MHz 8192 byte page size, 1 processor. CPU: EV45 (21064A) major=6 minor=2 OSF PAL rev: 0x100070002012d real memory = 132136960 (129040K bytes) avail memory = 122773504 (119896K bytes) Preloaded elf kernel "kernel" at 0xfffffc00005e6000. pci0: <PCI bus> on pcib0 sym0: <810a> irq 7 at device 5.0 on pci0 sym0: No NVRAM, ID 7, Fast-10, parity checking CACHE TEST FAILED: script execution failed. start=47c625c8, pc=47c625d4, end=47c625e8 sym0: CACHE INCORRECTLY CONFIGURED. device_probe_and_attach: sym0 attach returned 6 isp0: <Qlogic ISP 1020/1040 PCI SCSI Adapter> irq 1 at device 7.0 on pci0 isp0: interrupting at APECS irq 1 So, I do have an ncr810a chip after all embedded on this Alpha mainboard. My other PCI Alpha is on loan to another FreeBSD developer (of sigset fame ;-) so I can't check that one. After spending some time digging in my equipment stock I found a original NCR-built PCI card. Installing that gives me: sym1: <810a> irq 1 at device 7.0 on pci0 sym1: No NVRAM, ID 7, Fast-10, parity checking CACHE TEST FAILED: script execution failed. start=47c625c8, pc=47c625d4, end=47c625e8 sym1: CACHE INCORRECTLY CONFIGURED. device_probe_and_attach: sym1 attach returned 6 In both cases there are no devices connect to the SCSI buses that the sym driver controls. Cheers, Wilko -- | / o / / _ Arnhem, The Netherlands - Powered by FreeBSD - |/|/ / / /( (_) Bulte WWW : http://www.tcja.nl http://www.freebsd.org To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-scsi" in the body of the message
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?199911252040.VAA02381>