Dear Jeff,
The patch will increase waiting time in SCB initialization. It will
resolve the issue on STL2 motherboard. Pls apply.
--
Yours truly,
Louis Zhuang
---------------
Fault Injection Test Harness Project
BK tree: http://fault-injection.bkbits.net/linux-2.5
Home Page: http://sf.net/projects/fault-injection
You can import this changeset into BK by piping this whole message to:
'| bk receive [path to repository]' or apply the patch as usual.
===================================================================
[email protected], 2003-01-14 13:45:36+08:00, [email protected]
fix e100 initialization issue on STL2 motherboard
e100.h | 2 +-
1 files changed, 1 insertion(+), 1 deletion(-)
diff -Nru a/drivers/net/e100/e100.h b/drivers/net/e100/e100.h
--- a/drivers/net/e100/e100.h Tue Jan 14 13:46:26 2003
+++ b/drivers/net/e100/e100.h Tue Jan 14 13:46:26 2003
@@ -100,7 +100,7 @@
#define E100_MAX_NIC 16
-#define E100_MAX_SCB_WAIT 100 /* Max udelays in wait_scb */
+#define E100_MAX_SCB_WAIT 5000 /* Max udelays in wait_scb */
#define E100_MAX_CU_IDLE_WAIT 50 /* Max udelays in wait_cus_idle */
/* HWI feature related constant */
===================================================================
This BitKeeper patch contains the following changesets:
1.1019
## Wrapped with gzip_uu ##
begin 664 bkpatch4832
M'XL(`#*D(SX``\U4:VO;,!3]'/V*"_W6$OM>V7)20T;2!UMIRTH?;-^":LNU
M%ML:EM(7_O%3G)%22O9B@TG"$KK2T3WG'KP#-U:UZ:`R2VW9#GPPUJ6#4CXL
M`EL&NG&J"C)3^\BE,3X2EJ9687\ZY($8%IKYV(5T60GWJK7I@()HL^.>OJIT
M<'G\_N9L=LG89`*'I6SNU)5R,)DP9]I[6>5V*EU9F29PK6QLK9Q</=EMCG8<
MD?LN:!2A2#I*,!YU&>5$,B:5(X_'2<RLDTVEGJ:U;)S.3*M>,7@-%R%1[`'C
M*.G\;1&S(Z"`D/8!HQ`[email protected]>CE-$Z#E/WR@#>P1#9`?P=\D<L@P*
M_0B*_-.ZT4[+2C]+ITT#VMJE`K^XNC[C4!M7JO;6R#9GI\!%A)Q=O.C,AK_9
M&$.)[!U\N9/MLUZLY,Q;N5#6%*ZGDK=Z5>FP42Y<I==_@G)-+,(1CCGG48>C
M?2&Z0JJ"N,!1,LYX,8ZWR?@CU._%BCVUE5R4]$[:<F'EJW^3^@:V57DI?PKH
ML\:($W).'8G]9&TQ3F\,AK]@,/IO#+8NP$<8M@_]\(:YV%:+/_#>D1<-B)VL
MIYU<%;I1<.SAYN>SS_.KPX/YI]G)]4`@XB#<A7/Y",M<5?+)>A;P(+6;V^P6
;=L.7_U)6JFQAE_5D7,1"B$*P;QI*PE?V!```
`
end
On Tue, Jan 14, 2003 at 01:51:55PM +0800, Louis Zhuang wrote:
> The patch will increase waiting time in SCB initialization. It will
> resolve the issue on STL2 motherboard. Pls apply.
Sorry, not applied.
I was kinda hoping Scott would fix that up. It is a verified problem
(SMBus timeout, IIRC?), and this does indeed fix the problem.
However one should not udelay() that long -- especially in this case.
Since it is during init, sleep using schedule_timeout() for as long as
you want...
> begin 664 bkpatch4832
Please do not bother with sending BK patches like this. Either send
just the diff, as you did (thanks), or in addition send a URL from which
I may issue
bk pull $url
Regards,
Jeff
Jeff Garzik wrote:
> On Tue, Jan 14, 2003 at 01:51:55PM +0800, Louis Zhuang wrote:
> > The patch will increase waiting time in SCB
> initialization. It will
> > resolve the issue on STL2 motherboard. Pls apply.
>
> Sorry, not applied.
>
> I was kinda hoping Scott would fix that up. It is a verified
> problem (SMBus timeout, IIRC?), and this does indeed fix the problem.
Agreed, don't apply this patch - it's not the right fix.
The solution requires a specific TCO workaround on 82559 to
schedule_timeout after hard reset of the controller, and then follow
that with a wait for TCO traffic to quiesce. Patch forthcoming...
-scott