]> git.kernelconcepts.de Git - karo-tx-linux.git/commitdiff
s390/chsc: fix SEI usage
authorSebastian Ott <sebott@linux.vnet.ibm.com>
Tue, 15 Jan 2013 18:02:01 +0000 (19:02 +0100)
committerMartin Schwidefsky <schwidefsky@de.ibm.com>
Wed, 16 Jan 2013 14:57:54 +0000 (15:57 +0100)
cbc0dd1 "s390/pci: CHSC PCI support for error and availability events"
introduced a new SEI notification type as part of pci support.
The way SEI was called with nt2 and nt0 consecutive broke the nt0
stuff used for channel subsystem notifications.

The reason why this was broken with the mentioned patch is that you
cannot selectively disable type 0 notifications (so even when asked
for type 2 only, type 0 could be presented).

The way to do it is to tell SEI which types of notification you can
process and -this is the important part- look at the SEI result which
notification type you actually received.

Reviewed-by: Peter Oberparleiter <peter.oberparleiter@de.ibm.com>
Tested-by: Michael Holzheu <holzheu@linux.vnet.ibm.com>
Signed-off-by: Sebastian Ott <sebott@linux.vnet.ibm.com>
Signed-off-by: Martin Schwidefsky <schwidefsky@de.ibm.com>
drivers/s390/cio/chsc.c

index 68e80e2734a4be89763ca466a48621911570bce4..10729bbceceda4f49432e5eb647e04ae9d1399c4 100644 (file)
@@ -283,7 +283,7 @@ struct chsc_sei_nt2_area {
        u8  ccdf[PAGE_SIZE - 24 - 56];  /* content-code dependent field */
 } __packed;
 
-#define CHSC_SEI_NT0   0ULL
+#define CHSC_SEI_NT0   (1ULL << 63)
 #define CHSC_SEI_NT2   (1ULL << 61)
 
 struct chsc_sei {
@@ -291,7 +291,8 @@ struct chsc_sei {
        u32 reserved1;
        u64 ntsm;                       /* notification type mask */
        struct chsc_header response;
-       u32 reserved2;
+       u32 :24;
+       u8 nt;
        union {
                struct chsc_sei_nt0_area nt0_area;
                struct chsc_sei_nt2_area nt2_area;
@@ -496,17 +497,17 @@ static int __chsc_process_crw(struct chsc_sei *sei, u64 ntsm)
                                css_schedule_eval_all();
                        }
 
-                       switch (sei->ntsm) {
-                       case CHSC_SEI_NT0:
+                       switch (sei->nt) {
+                       case 0:
                                chsc_process_sei_nt0(&sei->u.nt0_area);
-                               return 1;
-                       case CHSC_SEI_NT2:
+                               break;
+                       case 2:
                                chsc_process_sei_nt2(&sei->u.nt2_area);
-                               return 1;
+                               break;
                        default:
-                               CIO_CRW_EVENT(2, "chsc: unhandled nt (nt=%08Lx)\n",
-                                             sei->ntsm);
-                               return 0;
+                               CIO_CRW_EVENT(2, "chsc: unhandled nt=%d\n",
+                                             sei->nt);
+                               break;
                        }
                } else {
                        CIO_CRW_EVENT(2, "chsc: sei failed (rc=%04x)\n",
@@ -537,15 +538,7 @@ static void chsc_process_crw(struct crw *crw0, struct crw *crw1, int overflow)
        sei = sei_page;
 
        CIO_TRACE_EVENT(2, "prcss");
-
-       /*
-        * The ntsm does not allow to select NT0 and NT2 together. We need to
-        * first check for NT2, than additionally for NT0...
-        */
-#ifdef CONFIG_PCI
-       if (!__chsc_process_crw(sei, CHSC_SEI_NT2))
-#endif
-               __chsc_process_crw(sei, CHSC_SEI_NT0);
+       __chsc_process_crw(sei, CHSC_SEI_NT0 | CHSC_SEI_NT2);
 }
 
 void chsc_chp_online(struct chp_id chpid)