Re: self-hosting on small-ish i386 boxen?

2019-11-18 Thread John D. Baker
On Mon, 18 Nov 2019, Robert Swindells wrote:

> "John D. Baker"  wrote:
> >The combination of LLVM 10.0 and GCC 8 on i386 makes self-hosting on at
> >least one i386 machine challenging.
> 
> Which GPU ?
> 
> If it isn't a radeon then you might want to try with MKLLVMRT=no.

It is a radeon but it's an older one that's not supported by DRMKMS.
(Before they were excluded in the code, they would exhibit almost-
black-on-black video in the framebuffer console, although X worked fine.)

For this (and some other) machine(s), the kernel configures the legacy
DRM driver (it would be nice if the legacy-supporting X driver could be
named differently so it wouldn't require a symlink dance after updating).

(The legacy DRM drivers have their own problems lately.  See:

  http://mail-index.NetBSD.org/current-users/2019/11/18/msg037067.html

for more.)

Linking gallium when building netbsd-9 (GCC 7) works.

-- 
|/"\ John D. Baker, KN5UKS   NetBSD Darwin/MacOS X
|\ / jdbaker[snail]consolidated[flyspeck]net  OpenBSDFreeBSD
| X  No HTML/proprietary data in email.   BSD just sits there and works!
|/ \ GPGkeyID:  D703 4A7E 479F 63F8 D3F4  BD99 9572 8F23 E4AD 1645


Re: self-hosting on small-ish i386 boxen?

2019-11-18 Thread Robert Swindells


"John D. Baker"  wrote:
>The combination of LLVM 10.0 and GCC 8 on i386 makes self-hosting on at
>least one i386 machine challenging.

Which GPU ?

If it isn't a radeon then you might want to try with MKLLVMRT=no.








self-hosting on small-ish i386 boxen?

2019-11-18 Thread John D. Baker
The combination of LLVM 10.0 and GCC 8 on i386 makes self-hosting on at
least one i386 machine challenging.

This is a Thinkpad T42 w/2GB RAM and 2GB swap partition.  It runs an
up-to-date 9.0_BETA (self-hosted build).

Up until the switch to GCC 8, it was able to build -current without any
problems.  After the switch to GCC 8, there seems to be quite the shortage
of memory.  Linking "gallium" is the most intensive operation.  I've
been getting:

[...]
/d0/build/current/tools/i386/lib/gcc/i486--netbsdelf/8.3.0/../../../../i486--netbsdelf/bin/ld:
 failed to set dynamic section sizes: memory exhausted
collect2: error: ld returned 1 exit status
*** [gallium_dri.so.0.link] Error code 1
nbmake[7]: stopped in /d0/nbsd/current/src/external/mit/xorg/lib/gallium
1 error
nbmake[7]: stopped in /d0/nbsd/current/src/external/mit/xorg/lib/gallium
*** [dependall] Error code 2
nbmake[6]: stopped in /d0/nbsd/current/src/external/mit/xorg/lib/gallium
1 error

nbmake[6]: stopped in /d0/nbsd/current/src/external/mit/xorg/lib/gallium
*** Failed target:  dependall-gallium
*** Failed command: _makedirtarget() { dir="$1"; shift; target="$1"; shift; 
case "${dir}" in /*) this="${dir}/"; real="${dir}" ;; .) 
this="external/mit/xorg/lib/"; 
real="/d0/nbsd/current/src/external/mit/xorg/lib" ;; *) 
this="external/mit/xorg/lib/${dir}/"; 
real="/d0/nbsd/current/src/external/mit/xorg/lib/${dir}" ;; esac; 
show=${this:-.}; echo "${target} ===> ${show%/}${1:+ (with: $@)}"; cd "${real}" 
&& /d0/build/current/tools/i386/bin/nbmake _THISDIR_="${this}" "$@" ${target}; 
}; _makedirtarget gallium dependall
*** Error code 2
Stop.


Adding additional swap via a swap file (up to 12GB now) has not worked.

Meanwhile, 4GB of RAM (w/4GB swap) on an amd64 host appears to be
sufficient for building -current (for now).

-- 
|/"\ John D. Baker, KN5UKS   NetBSD Darwin/MacOS X
|\ / jdbaker[snail]consolidated[flyspeck]net  OpenBSDFreeBSD
| X  No HTML/proprietary data in email.   BSD just sits there and works!
|/ \ GPGkeyID:  D703 4A7E 479F 63F8 D3F4  BD99 9572 8F23 E4AD 1645