Re: CVS commit: src/sys/arch/mips/mips
On 04/30/14 17:14, Izumi Tsutsui wrote: skrll@ wrote: I'm still seeing issues with my cobalt, but I'll keep hunting. Probably all mips port MD bus_dma.c (in arc, cobalt, emips, ews4800mips, hpcmips, mipsco, pmax, sgimips) need the similar fix as mips/bus_dma.c rev 1.28. yeah, they'll need that fix, but there are still cache aliasing bugs in the pmap, I think. (only algor, evbmips and sbmips use mips/bus_dma.c) --- Izumi Tsutsui
Re: CVS commit: src/sys/arch/mips/mips
skrll@ wrote: > yeah, they'll need that fix, but there are still cache aliasing bugs in > the pmap, I think. PR/45746 is still open, btw. --- Izumi Tsutsui
Re: CVS commit: src/sys/arch/mips/mips
Hello, On Wed, 30 Apr 2014 14:35:46 +0100 Nick Hudson wrote: > On 04/30/14 11:01, Michael wrote: > > Hello, > > > > On Wed, 23 Apr 2014 20:57:15 + > > "Nick Hudson" wrote: > > > >> Module Name: src > >> Committed By: skrll > >> Date: Wed Apr 23 20:57:15 UTC 2014 > >> > >> Modified Files: > >>src/sys/arch/mips/mips: pmap.c vm_machdep.c > > ... > >> Hopefully this addresses the instability reported in the following PRs: > >> > >> PR/44900 - R5000/Rm5200 mips ports are broken > >> PR/46170 - NetBSD/cobalt 6.0_BETA does not boot > >> PR/46890 - upcoming NetBSD 6.0 release is very unstable / unusable on > >> cobalt qube 2 > >> PR/48628 - cobalt and hpcmips ports are dead > > I can no longer reproduce the data corruption issues I've seen on sgimips > > in the last few years. > > Can you post your dmesg | grep ^cpu > > I'm still seeing issues with my cobalt, but I'll keep hunting. ... and just after sending the previous mail the data corruption issues popped up again. So they're not quite gone yet, just a whole lot less frequent. have fun Michael
Re: CVS commit: src/sys/arch/mips/mips
> they should also switch to the common mips bus space/dma files. - does it handle MIPS1? - BUS_DMASYNC_PREREAD seems incomplete --- Izumi Tsutsui
Re: CVS commit: src/sys/arch/mips/mips
Hello, On Wed, 30 Apr 2014 14:35:46 +0100 Nick Hudson wrote: > On 04/30/14 11:01, Michael wrote: > > On Wed, 23 Apr 2014 20:57:15 + > > "Nick Hudson" wrote: > > > >> Module Name: src > >> Committed By: skrll > >> Date: Wed Apr 23 20:57:15 UTC 2014 > >> > >> Modified Files: > >>src/sys/arch/mips/mips: pmap.c vm_machdep.c > > ... > >> Hopefully this addresses the instability reported in the following PRs: > >> > >> PR/44900 - R5000/Rm5200 mips ports are broken > >> PR/46170 - NetBSD/cobalt 6.0_BETA does not boot > >> PR/46890 - upcoming NetBSD 6.0 release is very unstable / unusable on > >> cobalt qube 2 > >> PR/48628 - cobalt and hpcmips ports are dead > > I can no longer reproduce the data corruption issues I've seen on sgimips > > in the last few years. > > Can you post your dmesg | grep ^cpu It's an O2: cpu0 at mainbus0: MIPS R5000 CPU (0x2321) Rev. 2.1 with built-in FPU Rev. 1.0 cpu0: 48 TLB entries, 1TB (40-bit) VAs, 64GB (36-bit) PAs, 16MB max page size cpu0: 32KB/32B 2-way set-associative L1 instruction cache cpu0: 32KB/32B 2-way set-associative write-back L1 data cache cpu0: 1024KB/32B direct-mapped write-through L2 unified cache > I'm still seeing issues with my cobalt, but I'll keep hunting. I cvs updated, built a new kernel and saw problems again. Older kernel ( sources from just after this commit) is ok. No clue why, I'll keep digging, might be local crap or PEBCAK. Also, what I've been messing with is both kernel and userland built as n32, guess I should try o32 too. With n32 I get the same compiler problems as on Loongson - SIGBUS from gmp whenever optimization is enabled. One built without .asm routines in gmp seems to work so far ( managed to build a working tcsh ) There's a possibly related issue on Loongson - nfs writes get corrupted in much the same way. No such problem with the working kernel on the O2 though. have fun Michael
Re: CVS commit: src/sys/arch/mips/mips
On Apr 30, 2014, at 9:14 AM, Izumi Tsutsui wrote: > skrll@ wrote: > >> I'm still seeing issues with my cobalt, but I'll keep hunting. > > Probably all mips port MD bus_dma.c > (in arc, cobalt, emips, ews4800mips, hpcmips, mipsco, pmax, sgimips) > need the similar fix as mips/bus_dma.c rev 1.28. > (only algor, evbmips and sbmips use mips/bus_dma.c) they should also switch to the common mips bus space/dma files.
Re: CVS commit: src/sys/arch/mips/mips
skrll@ wrote: > I'm still seeing issues with my cobalt, but I'll keep hunting. Probably all mips port MD bus_dma.c (in arc, cobalt, emips, ews4800mips, hpcmips, mipsco, pmax, sgimips) need the similar fix as mips/bus_dma.c rev 1.28. (only algor, evbmips and sbmips use mips/bus_dma.c) --- Izumi Tsutsui
Re: CVS commit: src/sys/arch/mips/mips
On 04/30/14 11:01, Michael wrote: Hello, On Wed, 23 Apr 2014 20:57:15 + "Nick Hudson" wrote: Module Name:src Committed By: skrll Date: Wed Apr 23 20:57:15 UTC 2014 Modified Files: src/sys/arch/mips/mips: pmap.c vm_machdep.c ... Hopefully this addresses the instability reported in the following PRs: PR/44900 - R5000/Rm5200 mips ports are broken PR/46170 - NetBSD/cobalt 6.0_BETA does not boot PR/46890 - upcoming NetBSD 6.0 release is very unstable / unusable on cobalt qube 2 PR/48628 - cobalt and hpcmips ports are dead I can no longer reproduce the data corruption issues I've seen on sgimips in the last few years. Can you post your dmesg | grep ^cpu I'm still seeing issues with my cobalt, but I'll keep hunting. Should we ever meet in real life, I owe you a beer or ten. Great, I'll remember :) have fun Michael Nick
Re: CVS commit: src/sys/arch/mips/mips
Hello, On Wed, 23 Apr 2014 20:57:15 + "Nick Hudson" wrote: > Module Name: src > Committed By: skrll > Date: Wed Apr 23 20:57:15 UTC 2014 > > Modified Files: > src/sys/arch/mips/mips: pmap.c vm_machdep.c ... > Hopefully this addresses the instability reported in the following PRs: > > PR/44900 - R5000/Rm5200 mips ports are broken > PR/46170 - NetBSD/cobalt 6.0_BETA does not boot > PR/46890 - upcoming NetBSD 6.0 release is very unstable / unusable on cobalt > qube 2 > PR/48628 - cobalt and hpcmips ports are dead I can no longer reproduce the data corruption issues I've seen on sgimips in the last few years. Should we ever meet in real life, I owe you a beer or ten. have fun Michael