Re: AMD Ryzen and NetBSD?

[Thor Lancelot Simon]

On Sun, Jul 02, 2017 at 10:57:20PM +0100, Patrick Welche wrote:
> On Fri, Jun 30, 2017 at 12:00:45PM -0400, Thor Lancelot Simon wrote:
> 
> I shoved a rather newer ST2000DM001-1CH164 in, which according to its
> marketing bumpf can manage "Max SustainableTransfer Rate 210MB/s"
> and not so bad:
> 
> # dd if=/dev/zero ibs=64k | progress -l 976751887b dd of=/dev/rdk15 obs=64
> k
>  99% |** |   465 GiB  116.74 MiB/s00:00 
> ETAd

This is already effectively double buffered, because of the way you used
"progress".  You could try using a larger blocksize for the reads from
/dev/zero (1m perhaps) and also for the writes to rdk15 - the kernel
will buffer up and dispatch the MAXPHYS sized I/Os.

To get 200MB out of that drive you likely need larger writes, which we
currently can't do.  It might perform slightly better through the
filesystem, though.

Thor

daily CVS update output

[NetBSD source update]

Updating src tree:
P src/etc/etc.evbarm/Makefile.inc
P src/games/fortune/datfiles/fortunes2
P src/lib/libpthread/TODO
P src/lib/libpthread/pthread.c
P src/lib/libpthread/pthread_attr.c
P src/lib/libpthread/pthread_attr_getguardsize.3
P src/lib/libpthread/pthread_int.h
P src/sys/arch/amd64/amd64/gdt.c
P src/sys/arch/amiga/amiga/trap.c
P src/sys/arch/arm/allwinner/awin_can.c
P src/sys/arch/arm/arm/syscall.c
P src/sys/arch/arm/arm/undefined.c
P src/sys/arch/arm/arm32/arm32_boot.c
P src/sys/arch/arm/arm32/arm32_machdep.c
P src/sys/arch/arm/arm32/fault.c
P src/sys/arch/arm/fdt/gic_fdt.c
P src/sys/arch/arm/samsung/exynos_i2c.c
P src/sys/arch/arm/samsung/exynos_pinctrl.c
P src/sys/arch/arm/sunxi/files.sunxi
P src/sys/arch/arm/sunxi/sun6i_a31_ccu.c
U src/sys/arch/arm/sunxi/sun6i_a31_gpio.c
P src/sys/arch/arm/sunxi/sun8i_h3_ccu.c
U src/sys/arch/arm/sunxi/sun8i_h3_gpio.c
U src/sys/arch/arm/sunxi/sunxi_gpio.c
U src/sys/arch/arm/sunxi/sunxi_gpio.h
U src/sys/arch/arm/sunxi/sunxi_rsb.c
U src/sys/arch/arm/sunxi/sunxi_rsb.h
P src/sys/arch/evbarm/conf/README.evbarm
P src/sys/arch/evbarm/conf/SUNXI
U src/sys/arch/evbarm/conf/SUNXI_INSTALL
P src/sys/arch/evbarm/conf/std.beagle
P src/sys/arch/evbarm/conf/std.exynos
P src/sys/arch/evbarm/conf/std.sunxi
P src/sys/arch/evbarm/conf/std.vexpress
P src/sys/arch/i386/i386/gdt.c
P src/sys/arch/i386/i386/mptramp.S
P src/sys/arch/i386/include/gdt.h
P src/sys/arch/i386/include/segments.h
P src/sys/arch/m68k/include/cpuframe.h
P src/sys/dev/fdt/fdt_pinctrl.c
P src/sys/dev/fdt/fdt_subr.c
P src/sys/dev/fdt/fdtvar.h
P src/sys/dev/ofw/ofw_subr.c
P src/sys/dev/ofw/openfirm.h
P src/sys/dev/pad/pad.c
P src/sys/kern/exec_subr.c
P src/sys/kern/kern_condvar.c
P src/sys/kern/kern_event.c
P src/sys/uvm/uvm_meter.c
P src/sys/uvm/uvm_param.h
P src/tests/lib/libpthread/t_join.c
P src/usr.bin/calendar/calendars/calendar.holiday

Updating xsrc tree:


Killing core files:



Updating release-6 src tree (netbsd-6):

Updating release-6 xsrc tree (netbsd-6):



Updating release-7 src tree (netbsd-7):

Updating release-7 xsrc tree (netbsd-7):



Updating release-8 src tree (netbsd-8):
U doc/CHANGES-8.0
P share/man/man4/tc.4
P sys/arch/alpha/tc/tcasic.c
P sys/arch/arm/allwinner/awin_can.c
P sys/arch/evbarm/conf/BCM5301X
P sys/arch/evbarm/conf/BCM56340
P sys/arch/evbarm/conf/BEAGLEBOARDXM
P sys/arch/evbarm/conf/BEAGLEBONE
P sys/arch/evbarm/conf/GEMINI
P sys/arch/evbarm/conf/GEMINI_MASTER
P sys/arch/evbarm/conf/GEMINI_SLAVE
P sys/arch/evbarm/conf/GOLDENGATE
P sys/arch/evbarm/conf/N900
P sys/arch/evbarm/conf/PANDABOARD
P sys/arch/pmax/tc/tcbus.c
P sys/arch/vax/conf/files.vax
P sys/arch/vax/vsa/tc_vsbus.c
P sys/dev/sdmmc/ld_sdmmc.c
P sys/dev/sdmmc/sdmmc_mem.c
P sys/dev/sdmmc/sdmmcreg.h
P sys/dev/sdmmc/sdmmcvar.h
P sys/dev/tc/tc.c
P sys/dev/tc/tcvar.h
P sys/fs/udf/udf_vfsops.c
P sys/miscfs/specfs/spec_vnops.c
P sys/net/if.c
P sys/net/if.h
P sys/netinet/if_arp.c
P sys/netinet6/ip6_output.c

Updating release-8 xsrc tree (netbsd-8):




Updating file list:
-rw-rw-r--  1 srcmastr  netbsd  56927248 Jul  3 03:11 ls-lRA.gz

Re: Dark screen with nouveau

[Chavdar Ivanov]

s/earlier/elsewhere/ - the problem could be in another place, introduced
later. So no idea.

Chavdar

On Fri, 30 Jun 2017 at 11:51 Chavdar Ivanov  wrote:

> Thanks for the suggestion, but it appears the problem was earlier. I built
> an up-to-date kernel with the four files downgraded to the versions prior
> to your change, but the result was the same - the system working fine but
> with a black screen.  Xorg also starts, judging by the processes present,
> but does not alight the screen.  This is with the following versions:
> ---
> ➜  conf # head -1 /usr/src/sys/external/bsd/drm2/include/drm/drmfb.h
> /*$NetBSD: drmfb.h,v 1.1 2015/03/05 17:50:41 riastradh Exp $
>  */
> ➜  conf # strings /netbsd | grep \$NetBSD | egrep
> intelfb.c\|nouveaufb.c\|drmfb.c
> $NetBSD: drmfb.c,v 1.2 2015/11/09 23:11:18 jmcneill Exp $
> $NetBSD: intelfb.c,v 1.12 2015/03/05 17:56:39 riastradh Exp $
> $NetBSD: nouveaufb.c,v 1.3 2015/10/17 12:02:44 jmcneill Exp $
>
> Regards,
>
> Chavdar
>
>
> On Thu, 29 Jun 2017 at 03:36  wrote:
>
>> On Wed, Jun 28, 2017 at 02:34:55PM +, Chavdar Ivanov wrote:
>>
>> > It would be a bit tedious to do bisecting from 20th of November until
>> now,
>> > so I hoped someone would remember some change for this.
>> >
>>
>> Most likely I broke it with this change:
>> https://v4.freshbsd.org/commit/netbsd/src/dY8Umnn16RPolJxz
>>
>

Re: AMD Ryzen and NetBSD?

[Patrick Welche]

On Fri, Jun 30, 2017 at 12:00:45PM -0400, Thor Lancelot Simon wrote:
> On Fri, Jun 30, 2017 at 04:55:40PM +0100, Patrick Welche wrote:
> > On Fri, Jun 30, 2017 at 03:41:39PM +, m...@netbsd.org wrote:
> > > On Fri, Jun 30, 2017 at 09:25:19AM +0100, Patrick Welche wrote:
> > > > I'm happily running -current on a Ryzen 7 1700 / Asus PRIME X370-PRO.
> > > > I wonder about its i/o though: I remember dd'ing zeros over old disks, 
> > > > and
> > > > thought "this is slow", so I dd'd over 3 disks at once, with no change
> > > > in individual "progress" bit rate, i.e. 3 times the bandwidth...
> > > 
> > > Are you specifying block size? dd is terribly slow without it.
> > 
> > I was using 64k - essentially the example in the "progress" man page.
> 
> Latency.  You want a double-buffered pipeline.

I shoved a rather newer ST2000DM001-1CH164 in, which according to its
marketing bumpf can manage "Max SustainableTransfer Rate 210MB/s"
and not so bad:

# dd if=/dev/zero ibs=64k | progress -l 976751887b dd of=/dev/rdk15 obs=64
k
 99% |** |   465 GiB  116.74 MiB/s00:00 ETAd
d: /dev/rdk15: short write on character device
dd: /dev/rdk15: end of device
976752000+0 records in
7630874+1 records out
500096966144 bytes transferred in 4085.217 secs (122416255 bytes/sec)
100% |***|   465 GiB  116.74 MiB/s--:-- ETA


and trying the double buffering (is this the right idea?)


quantz# dd if=/dev/zero ibs=64k | dd bs=64k | progress -l 976751887b dd of=/dev/
rdk15 obs=64k   
 99% |** |   465 GiB  116.04 MiB/s00:00 ETAd
d: /dev/rdk15: short write on character device
dd: /dev/rdk15: end of device
976752000+0 records in
7630874+1 records out
500096966144 bytes transferred in 4109.835 secs (121682979 bytes/sec)
100% |***|   465 GiB  116.04 MiB/s--:-- ETA

Re: hdaudio1: RIRB timeout in 8.0_BETA

[Kamil Rytarowski]

On 02.07.2017 14:10, g...@duzan.org wrote:
> "Nathanial Sloss"  wrote:
> => Hi,
> =>
> => Just wondering if your NetBSD-8 snapshot was from sources after 20170629.
> =>
> => Best regards,
> =>
> => Nat
> 
>I believe so, yes. I maintain an rsync copy of the CVS repo, which
> would have been updated at 2017-06-30 05:05 EDT, and my working tree
> update from it completed at 2017-06-30 08:17 EDT. I'll do another
> update/build/upgrade in case I got the tree in a weird state.
> 
>Thanks.
> 
>  Gary Duzan
> 
> 
> 

Reproduced here the same issue with NetBSD-current July 1st sources.



signature.asc
Description: OpenPGP digital signature

Re: hdaudio1: RIRB timeout in 8.0_BETA

[gary]

"Nathanial Sloss"  wrote:
=> Hi,
=>
=> Just wondering if your NetBSD-8 snapshot was from sources after 20170629.
=>
=> Best regards,
=>
=> Nat

   I believe so, yes. I maintain an rsync copy of the CVS repo, which
would have been updated at 2017-06-30 05:05 EDT, and my working tree
update from it completed at 2017-06-30 08:17 EDT. I'll do another
update/build/upgrade in case I got the tree in a weird state.

   Thanks.

 Gary Duzan

Re: hdaudio1: RIRB timeout in 8.0_BETA

[Nathanial Sloss]

Hi,

Just wondering if your NetBSD-8 snapshot was from sources after 20170629.

Best regards,

Nat