[9fans] 9pi and upas/send question

[9fans] 9pi and upas/send question

[Francisco J Ballesteros]

To anyone with a pi out there,

can you try to do an upas send on the pi (just send a mail)
and let me know if it worked fine there?

thanks

Re: [9fans] 9pi and upas/send question

[Richard Miller]

> can you try to do an upas send on the pi (just send a mail)

I do it all the time.  (This message for example.)

Re: [9fans] 9pi and upas/send question

[Fco. J. Ballesteros]

[-- Attachment #1: Type: text/plain, Size: 81 bytes --]

Thanks.
It seems our send is broken just on the pi and I'm trying to
see why.

[-- Attachment #2: Type: message/rfc822, Size: 4813 bytes --]

From: Richard Miller <9fans@hamnavoe.com>
To: 9fans@9fans.net
Subject: Re: [9fans] 9pi and upas/send question
Date: Tue, 21 May 2013 11:36:26 +0100
Message-ID: <e432032cd747c50f5c16f12df5477422@hamnavoe.com>

> can you try to do an upas send on the pi (just send a mail)

I do it all the time.  (This message for example.)

Re: [9fans] 9pi and upas/send question

[Richard Miller]

> It seems our send is broken just on the pi and I'm trying to
> see why.

I did have an intermittent tcp problem (which I still haven't debugged)
causing upas/send to hang, which I worked around by setting mtu to 1496.
I think I got that on x86 as well as the pi, but I'm not certain.

Can you define "broken" more specifically?

Re: [9fans] 9pi and upas/send question

[Fco. J. Ballesteros]

[-- Attachment #1: Type: text/plain, Size: 201 bytes --]

It breaks trying to use a biobuf and that
biobuf* has -1 as the value of the pointer.

We usually pull the source and then recompile all the
binaries.

Perhaps it's something we made.

thanks

[-- Attachment #2: Type: message/rfc822, Size: 5031 bytes --]

From: Richard Miller <9fans@hamnavoe.com>
To: 9fans@9fans.net
Subject: Re: [9fans] 9pi and upas/send question
Date: Tue, 21 May 2013 11:47:43 +0100
Message-ID: <8ff950de68e7efe433f04afe841c0c93@hamnavoe.com>

> It seems our send is broken just on the pi and I'm trying to
> see why.

I did have an intermittent tcp problem (which I still haven't debugged)
causing upas/send to hang, which I worked around by setting mtu to 1496.
I think I got that on x86 as well as the pi, but I'm not certain.

Can you define "broken" more specifically?