[9fans] what arch

[9fans] what arch

[ron minnich]

somebody referred me to the discussion.

Sometimes we found people wanted to build on their existing OS (Linux,
OSX, whatever) in a cross-build way, and, further, didn't want to do
that in a VM, because they had tools they liked.

github.com/rminnich/NxM is the last snapshot of the Sandia/BL fork,
and it has scripts and instructions to crossbuild it all on Linux.
It's not elegant but it works. At the time, we used Gerrit and Jenkins
for our control and validation. For each commit, gerrit would kick off
a jenkins run, which would do the full build from scratch, boot in
qemu, and run a set of regression tests. Gerrit would -1 the patch if
the jenkins pass did not work.

Full build, starting from nothing, of tools, libs, bin, kernels, was
about two minutes on Linux. If you added gs into the mix, it was more
like 4 minutes IIRC. Ran fine on amd64.

Also, from the Akaros side, we've extended the Plan 9 MSI code to
support MSI-X and tested it on real hardware; works fine. This allowed
us to stop worrying about ACPI; you don't need it if you do MSI-X.
Relief! And MSI-X is in some ways easier than MSI.

There's also a first cut at a vm device. I realize there's no love
lost for Akaros in this group, but you're welcome to the code. I've
also started porting oprofile into Akaros in a way that would allow
its use in Plan 9. We need to substantially improve the TCP stack
(which comes from Plan 9).

One suggestion I'd like to float here: the LPL is a problem for both
BSD and GPL worlds (see Theo's notes from 2003 on that issue). It
might be useful for new from-scratch software to be released under
3-clause BSD; or under the Inferno license for that matter. In other
words, if you don't have to use the LPL, use 3-clause BSD instead. One
person has already very kindly allowed us to use their drivers in
Akaros with the LPL replaced with 3-clause BSD.

Anyway, congrats on the GSOC and good luck.

ron

Re: [9fans] what arch

[erik quanstrom]

> Also, from the Akaros side, we've extended the Plan 9 MSI code to
> support MSI-X and tested it on real hardware; works fine. This allowed
> us to stop worrying about ACPI; you don't need it if you do MSI-X.
> Relief! And MSI-X is in some ways easier than MSI.

i don't see how this could be the case for some pretty common hardware.
for example, i have yet to see a usb controller that will do either
msi or msi-x.

for example, on this machine (random intel)

	; pci |grep usb
	0.26.0:	usb  0c.03.20 8086/1c2d  11 0:fbc23000 1024
	0.29.0:	usb  0c.03.20 8086/1c26   5 0:fbc22000 1024
	; pci -vc 0.26.0 0.29.0
	0.26.0: pwrmgmt/01 dbgprt/0a advfeat/13
	0.29.0: pwrmgmt/01 dbgprt/0a advfeat/13

nope.  it's i/o apic interrupts for me!  unless there's some magic way
to convert i/o apic into msi-x that i don't know about.  hoping this
is the case....

- erik

Re: [9fans] what arch

[erik quanstrom]

actually, on this machine, about half the interrupt
sources don't do msi, since msi is the default irq type:

; grep ioapic /dev/irqalloc
#vec.mach	irq	count	cycles	type	name
       65.0           1                59967            793975781 ioapic   i8042
       65.7           9                    0                    0 ioapic   acpi
       66.1           4                66350           8910023002 ioapic   COM1
       66.3           5             18364273          98144708590 ioapic   usbehci
       66.6          11             18182865         162682346167 ioapic   usbehci
; grep msi /dev/irqalloc
       65.1           7                    0                    0 msi      ether1
       65.2           5                    0                    0 msi      ether3
       65.4          11              6002102          30018646458 msi      ether0
       65.5          11                    0                    0 msi      ether2
       66.4          11                   14               248036 msi      sdE (ahci)

- erik

Re: [9fans] what arch

[Bakul Shah]

On Fri, 09 May 2014 13:37:04 PDT ron minnich <rminnich@gmail.com> wrote:
> somebody referred me to the discussion.
>
> Sometimes we found people wanted to build on their existing OS (Linux,
> OSX, whatever) in a cross-build way, and, further, didn't want to do
> that in a VM, because they had tools they liked.
>
> github.com/rminnich/NxM is the last snapshot of the Sandia/BL fork,
> and it has scripts and instructions to crossbuild it all on Linux.
> It's not elegant but it works. At the time, we used Gerrit and Jenkins
> for our control and validation. For each commit, gerrit would kick off
> a jenkins run, which would do the full build from scratch, boot in
> qemu, and run a set of regression tests. Gerrit would -1 the patch if
> the jenkins pass did not work.
>
> Full build, starting from nothing, of tools, libs, bin, kernels, was
> about two minutes on Linux. If you added gs into the mix, it was more
> like 4 minutes IIRC. Ran fine on amd64.

Seems very slow : )

Full plan9 *native* build of the kernel, libs and bin on a
/RapsberryPi/ is about 4 minutes Crossbuilding i386 kernel on
it takes about 3 minutes (I haven't tried a full crossbuild).
Building the 9pi kernel under 9vx takes about 11 seconds on my
MBP @ home. Don't recall the full build time.

For comparison, a native Linux kernel build on RPi takes over
10 hours.

For another comparison, RPi seems about 16-20 time slower
compared to a MBP.

> One suggestion I'd like to float here: the LPL is a problem for both
> BSD and GPL worlds (see Theo's notes from 2003 on that issue). It
> might be useful for new from-scratch software to be released under
> 3-clause BSD; or under the Inferno license for that matter. In other
> words, if you don't have to use the LPL, use 3-clause BSD instead. One
> person has already very kindly allowed us to use their drivers in
> Akaros with the LPL replaced with 3-clause BSD.

Agree.

Re: [9fans] what arch

[ron minnich]

On Fri, May 9, 2014 at 1:51 PM, Bakul Shah <bakul@bitblocks.com> wrote:

> Full plan9 *native* build of the kernel, libs and bin on a
> /RapsberryPi/ is about 4 minutes

GOOD. Why not have a web page? The great plan 9 build shootout. Nobody
would be happier than me if Linux always lost.

ron

Re: [9fans] what arch

[Charles Forsyth]

[-- Attachment #1: Type: text/plain, Size: 133 bytes --]

On 9 May 2014 21:37, ron minnich <rminnich@gmail.com> wrote:

> under the Inferno license for that matter


that's usually MIT

[-- Attachment #2: Type: text/html, Size: 423 bytes --]

Re: [9fans] what arch

[ron minnich]

On Fri, May 9, 2014 at 2:46 PM, Charles Forsyth
<charles.forsyth@gmail.com> wrote:
>
> On 9 May 2014 21:37, ron minnich <rminnich@gmail.com> wrote:
>>
>> under the Inferno license for that matter
>
>
> that's usually MIT

Yeah. Either 3 clause BSD or MIT work. Many others don't. So whichever
one of those floats your boat, those of you have authored new software
might consider relicensing under one of them. Again, just a
suggestion.

ron