Subject: datapoint: build of -current on Compaq DS10 (Alpha) = ~6.5 hours.
To: None <netbsd-users@netbsd.org>
From: Johan A.van Zanten <johan@giantfoo.org>
List: netbsd-users
Date: 10/29/2005 14:15:43
Just wanted to share some build times -- this is the fastest machine i
have at home.

Build started at:  Sat Oct 29 01:33:25 CDT 2005
Build finished at: Sat Oct 29 07:28:41 CDT 2005
===> Successful make build
===> build.sh started: Sat Oct 29 01:33:22 CDT 2005
===> build.sh ended:   Sat Oct 29 07:28:41 CDT 2005
===> Summary of results:
         build.sh command: ./build.sh -o -m sparc64 -U -u -T /tew/netbsd-src/NetBSD-current/src/tooldir.NetBSD-2.0.2_STABLE-alpha build

Machine info:

NetBSD 2.0.2_STABLE (PARATHA) #1: Sun Apr 17 18:26:19 CDT 2005
        johan@sarasvati:/local/NetBSD/src/NetBSD-2.0/src/sys/arch/alpha/compile/
obj.alpha/PARATHA
COMPAQ AlphaServer DS10 466 MHz, s/n r020dqmz00
8192 byte page size, 1 processor.
total memory = 1024 MB


 This was a complete build of -current, excluding X, and the GENERIC
kernel.  (I think the GENERIC kernel took about 30 minutes to build.) I
believe the tool chain was already built.

What's the build time like on a fast Intel-based machine now days?

 If i remember correctly, the last time i noticed the duration of a
complete build was probably 1.6 and that was about 8 hours on the same
machine.  It's interesting to me the build has become faster.  (I'm pretty
sure that 8 hour build of 1.6.x also did not include X.)

 The device getting all of the I/O during the build is:

   wd0 at atabus0 drive 0: <ST3160021A>

 I've seen some performance problems with that drive on this system under
NetBSD 2.0.2 (that myself and others have reported), so this may have
affected the build time.

For the curious who may not have seen the message to port-alpha, this is
what the kernel always reports just as it starts to use wd0:

root on sd0a dumps on sd0b
root file system type: ffs
aceride0:0:0: lost interrupt
        type: ata tc_bcount: 2048 tc_skip: 0
aceride0:0:0: bus-master DMA error: missing interrupt, status=0x21
wd0: transfer error, downgrading to Ultra-DMA mode 1
wd0(aceride0:0:0): using PIO mode 4, Ultra-DMA mode 1 (using DMA data transfers)
wd0h: DMA error reading fsbn 312168096 of 312168096-312168099 (wd0 bn 312168096;
 cn 309690 tn 9 sn 9), retrying
aceride0:0:0: lost interrupt
        type: ata tc_bcount: 2048 tc_skip: 0
aceride0:0:0: bus-master DMA error: missing interrupt, status=0x21
wd0: transfer error, downgrading to DMA mode 2
wd0(aceride0:0:0): using PIO mode 4, DMA mode 2 (using DMA data transfers)
wd0h: DMA error reading fsbn 312168096 of 312168096-312168099 (wd0 bn 312168096;
 cn 309690 tn 9 sn 9), retrying
aceride0:0:0: lost interrupt
        type: ata tc_bcount: 2048 tc_skip: 0
aceride0:0:0: bus-master DMA error: missing interrupt, status=0x21
wd0: transfer error, downgrading to PIO mode 4
wd0(aceride0:0:0): using PIO mode 4
wd0h: DMA error reading fsbn 312168096 of 312168096-312168099 (wd0 bn 312168096;
 cn 309690 tn 9 sn 9), retrying
wd0: soft error (corrected)

(I've tried different, supposedly better IDE cables.)

 -johan