s390 lcs incorrect test
While debugging why our LCS emulator is having some problems I noticed the following weirdness in drivers/s390/net/lcs.c routine lcs_irq:
lcs.c renamed from lcs.c.orig +1 -1
@@ -1354,7 +1354,7 @@ index = (struct ccw1 *)__va ((addr_t ) irb->scsw.cpa) - channel->ccws;if ((irb->scsw.actl &SCSW_ACTL_SUSPENDED ) || - (irb->scsw.cstat |SCHN_STAT_PCI )) + (irb->scsw.cstat &SCHN_STAT_PCI ))/* Bloody io subsystem tells us lies about cpa... */ index = (index -1 ) & (LCS_NUM_BUFFS -1 );while (channel->io_idx != index) {
The `if' statement is always true since SCHN_STAT_PCI is defined as 0x80. Don't know if this has anything to do with our LCS problems but thought I would pass it on.
Greg Smith
I agree , it's really weird ...
Thank you Greg for the patch ...
Mit freundlichen Grüssen / Best regards
Frank Pavlic
Linux for eServer Development
Schoenaicher Str. 220, 71032 Boeblingen
Phone: ext. +49-(0)7031/16-2463, int. *120-2463 mailto: [email protected]
Greg Smith <[email protected]>
27.04.2006 02:05
To
[email protected]
cc
[email protected], Frank Pavlic/Germany/IBM@IBMDE Subject
s390 lcs incorrect test
While debugging why our LCS emulator is having some problems I noticed the following weirdness in drivers/s390/net/lcs.c routine lcs_irq:
lcs.c 2006-04-26 19:56:45.000000000 -0400 renamed from lcs.c.orig 2006-04-24 16:20:24.000000000 -0400 +0 -0
@@ -1354,7 +1354,7 @@ index = (struct ccw1 *) __va((addr_t)
irb->scsw.cpa)
- channel->ccws;
if ((irb->scsw.actl &SCSW_ACTL_SUSPENDED) ||
- (irb->scsw.cstat | SCHN_STAT_PCI))
+ (irb->scsw.cstat & SCHN_STAT_PCI))
/* Bloody io subsystem
tells us lies about cpa... */
index = (index - 1) &
(LCS_NUM_BUFFS - 1);
while (channel->io_idx != index) {The `if' statement is always true since SCHN_STAT_PCI is defined as 0x80. Don't know if this has anything to do with our LCS problems but thought I would pass it on.
Greg Smith
I have discovered today our bug regarding lcs emulation and everything seems to work fine even with the incorrect test. Which leads me to think that the lcs code has been crafted to withstand the incorrect test. Meaning that correcting the test may have implications.
Greg Smith
Quoted 3 lines · Frank Pavlic
I agree , it's really weird ...
Thank you Greg for the patch ...