Re: [patch 1/2] m68k: Atari EtherNAT (SMC91C111) driver (updated)

[Michael Schmitz]

Hi Geert,


drivers/net/atari_91C111.c: In function 'smc_rcv':
drivers/net/atari_91C111.c:584: error: implicit declaration of function
'readsw'


Odd - that should have been added by my previous io.h patch (the one that was
too chatty in the comments, around July 20).


Ah, apparently I didn't apply the last hunk. Will fix.

(That's what happens when manually applying whitespace-challenged
patches :-)


I know - that's why I sent these one as attachments :-)


Duh - now where did that one come from? Seems someoene removed irq_flags from
the platform data struct between 2.6.26 and 2.6.28 ... I'll check the driver
changelog. Looks like I should work from a fresh tree anyway.


commit d280eadc4fba0bf99fb1c3b60e8c5e007f7da02c
Author: Eric Miao <[EMAIL PROTECTED]>
Date:   Fri Jun 6 17:13:02 2008 +0800

   [NET] smc91x: remove "irq_flags" from "struct smc91x_platdata"


Thanks. Seems I can just drop the irq_flags setting altogetheras well.


BTW, can you please run scripts/checkpatch.pl on your patches?

| total: 63 errors, 87 warnings, 2621 lines checked


Duh again ... half the errors were in the original smc91x.c but I'll clean up
my mess now.


Can they be merged?


Tough call - at this point, the Atari driver requires special case handling for 
setting up the timer interrupt (i.e., I can hardwire timer D as interrupt but I 
still need to program the timer for the first instance it gets used. May need 
to go into a separate module).


When I get the level 6 interrupt to work, it'll be so much easier to merge. I'll 
get to it :-)


Michael


--
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]

Re: [buildd] releasing lenny

[Stephen R Marenka]

On Fri, Sep 05, 2008 at 07:09:20AM +0200, Michael Schmitz wrote:
> Hi,
>
>> No, no binaries from lenny will be going back into unstable to make
>> sure. If it becomes an issue, it isn't very difficult to resign the
>> changes files after the fact.
>
> If you say so - no further objections from my part.
>
> Regarding archive administration - I suppose Stephen has access to that 
> as well? I would not mind trying my hand at it either if more backup is 
> needed.

At the moment, all I have access to is buildd_m68k. I'll happily
share that, send me an ssh key. :)

We've also got to find changes for the list at 
. For whatever
reason they are keeping us from syncing the archive. Otherwise we'll
have to rebuild them. I'm guessing they aren't on merkel, but I haven't
looked yet. At least some of them should be able to come out of the
buildd logs too. Anyone? Anyone?

Peace,

Stephen

-- 
Stephen R. Marenka If life's not fun, you're not doing it right!
<[EMAIL PROTECTED]>


signature.asc
Description: Digital signature

Re: [patch 3/2] m68k: Atari EtherNAT - add writew_be for data push

[Frank P. Szymanski]

Hi Michael,
> Comments welcome (is there anyone on this list with access to an
> EtherNAT card,
> besides cts and myself?).

Here (but I guess you already knew that?).
> Regarding performance afforded by having a 100 Mbit PHY instead of a
> 10 Mbit
> one:
>
> EtherNAT: 210 KB/s out, 140 KB/s in (scp of a 35 MB vmlinux file)
> EtherNEC: 175 KB/s out, 128 KB/s in (same file)
>
> Hardly worth it, eh?

Depends, with MiNT speed differences are more impressive:

Falcon/AB040 or /ct60 with EtherNEC:
80k/sec Internet or 350k/sec local ftp server

Falcon/ct60 with EtherNAT:
500k/sec Internet or 1.200k/sec local ftp server

Perhaps there is room for optimization?

Regards,

Frank


-- 
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]

Re: 2008 M68k Porter Meeting

[Riccardo]

Hey,

nice you had your meeting. I tried all three days to log into 
#debian-68k on freenode, but it was always empty.


good to here that there are good news...


Riccardo

On Wednesday, September 3, 2008, at 12:42  PM, Christian T. Steigies 
wrote:


The meeting is over, the weather was even nicer than promised (its 
back to

normal now...) and it seems everybody got back home safely.




--
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]

Re: [patch 3/2] m68k: Atari EtherNAT - add writew_be for data push

[Michael Schmitz]

Hi Frank,


Comments welcome (is there anyone on this list with access to an
EtherNAT card,
besides cts and myself?).


Here (but I guess you already knew that?).


Sure - I wasn't sure whether you still followed the list though.


Regarding performance afforded by having a 100 Mbit PHY instead of a
10 Mbit
one:

EtherNAT: 210 KB/s out, 140 KB/s in (scp of a 35 MB vmlinux file)
EtherNEC: 175 KB/s out, 128 KB/s in (same file)

Hardly worth it, eh?


Depends, with MiNT speed differences are more impressive:

Falcon/AB040 or /ct60 with EtherNEC:
80k/sec Internet or 350k/sec local ftp server


I guess I will have to switch tp FTP to eliminate the ssh overhead. I cannot 
seem to reproduce my receive speed anyway, the tests were done with both cards 
running (different subnets though) so who knows what happened.



Falcon/ct60 with EtherNAT:
500k/sec Internet or 1.200k/sec local ftp server

Perhaps there is room for optimization?


Using interrupts on the EtherNAT would be a good starting point. I get hundreds 
of RX overrun kernel messages, that's another bottleneck.


I'll have to clean up the code quite a bit to make it acceptable before I'll 
start on interrupts. Still have to test whether it all works when compiled in, 
too.


Is there a MiNT driver for the USB port on the EtherNAT as well? That might be 
useful to tackle next.


Michael


--
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]

Re: [buildd] releasing lenny

[Michael Schmitz]

Hi,


Regarding archive administration - I suppose Stephen has access to that
as well? I would not mind trying my hand at it either if more backup is
needed.


At the moment, all I have access to is buildd_m68k. I'll happily
share that, send me an ssh key. :)


I'll send the keys q650 and hobbes used so far, is that what you mean?


We've also got to find changes for the list at
. For whatever
reason they are keeping us from syncing the archive. Otherwise we'll
have to rebuild them. I'm guessing they aren't on merkel, but I haven't
looked yet. At least some of them should be able to come out of the
buildd logs too. Anyone? Anyone?


I'll see what I can find in the log archive on buildd.d.o. Not to mention my 
changes archive :-)


Michael


--
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]

Re: [buildd] releasing lenny

[Michael Schmitz]

Hi,


We've also got to find changes for the list at
. For whatever
reason they are keeping us from syncing the archive. Otherwise we'll
have to rebuild them. I'm guessing they aren't on merkel, but I haven't
looked yet. At least some of them should be able to come out of the
buildd logs too. Anyone? Anyone?


I have these (or more precisely, I either have copies of the signed files, or 
unsigned ones that I could re-sign):


arc_5.21m-1_m68k.changes
cabber_0.4.0-test5-2.1_m68k.changes
clif_0.93-8_m68k.changes
codegroup_19981025-3_m68k.changes
cookietool_2.5-2_m68k.changes
dangen_0.3-2_m68k.changes
elfsign_0.2.2-2_m68k.changes
firestarter_1.0.3-6_m68k.changes
fl-cow_0.6-1_m68k.changes
francine_0.99.8orig-6_m68k.changes
giftoxic_0.0.10-4_m68k.changes
gspot_0.1.5-1_m68k.changes
hrd_0.1-3_m68k.changes
hsetroot_1.0.2-1_m68k.changes
icemc_0.2.4-3_m68k.changes
jabber-aim_20040131b-2_m68k.changes
keynote_2.3-11_m68k.changes
ladcca_0.4.0-6_m68k.changes
libexif-gtk_0.3.5-3_m68k.changes
libgnome2-print-perl_0.95.1-1_m68k.changes
libmail-pop3client-perl_2.17-1_m68k.changes
libmimedir_0.4-4_m68k.changes
libnasl_2.2.10-1_m68k.changes
libopengl-ruby_0.32f-2_m68k.changes
lookup_1.08b-10_m68k.changes
lsnipes_0.9.4-6_m68k.changes
lv_4.51-2_m68k.changes
makejvf_1.1a+0-2_m68k.changes
mdetect_0.5.2.1_m68k.changes
minimuf_3.5-1_m68k.changes
modemu_0.0.1-10_m68k.changes
naim_0.11.8-1_m68k.changes
network-manager-vpnc_0.6.4svn2806-1_m68k.changes
ninvaders_0.1.1-2_m68k.changes
nyquist_2.29-6_m68k.changes
photopc_3.05-6_m68k.changes
portmidi_20041117-1_m68k.changes
potrace_1.8-1_m68k.changes
qsf_1.2.7-1_m68k.changes
safecat_1.12-1_m68k.changes
sdparm_1.02-1_m68k.changes
sm_2.1-0-1_m68k.changes
ssh-askpass-fullscreen_0.3-2_m68k.changes
stax_1.0-13_m68k.changes
sumika_0.11-2.1_m68k.changes
tuxpuck_0.8.2-2.1_m68k.changes
udo_6.4.1-1_m68k.changes
vdkxdb2_2.4.0-3_m68k.changes
vftool_2.0alpha-3_m68k.changes
viruskiller_1.0-1-1_m68k.changes
w3c-libwww_5.4.0-11_m68k.changes
wp2x_2.5-mhi-9_m68k.changes
xblast-tnt_2.10.4-3_m68k.changes
xkbsel_0.13-13.3_m68k.changes
xmove_2.0beta2-8_m68k.changes

About a quarter of them... do you need them all signed?

More might be lurking in my mail (above are all from hobbes, q650, kullervo and 
crest and were automatically extracted from log mails, but that may have failed 
for log mails from other hosts). I'll dig some more.


Michael


--
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]