On Tue, 2007-01-16 at 10:06 +0200, Michael S. Tsirkin wrote:
> > Quoting Steve Wise <[EMAIL PROTECTED]>:
> > Subject: [PATCH RFC 16/21] ofed_1_2 Backport
> > infinband/hw/cxgb3/iwch_cm.[ch] to sles9sp3
> >
> >
> > Backport infinband/hw/cxgb3/iwch_cm.[ch] to sles9sp3
> >
> > Signed-off-by: Steve Wise <[EMAIL PROTECTED]>
> > ---
> >
> > .../2.6.5_sles9_sp3/iwch_cm_to_2_6_5-7_244.patch | 35
> > +++++++++++++++++++++++
> > 1 files changed, 35 insertions(+), 0 deletions(-)
> >
> > diff --git
> > a/kernel_patches/backport/2.6.5_sles9_sp3/iwch_cm_to_2_6_5-7_244.patch
> > b/kernel_patches/backport/2.6.5_sles9_sp3/iwch_cm_to_2_6_5-7_244.patch
> > new file mode 100644
> > index 0000000..af468f7
> > --- /dev/null
> > +++ b/kernel_patches/backport/2.6.5_sles9_sp3/iwch_cm_to_2_6_5-7_244.patch
> > @@ -0,0 +1,35 @@
> > +diff --git a/drivers/infiniband/hw/cxgb3/iwch_cm.c
> > b/drivers/infiniband/hw/cxgb3/iwch_cm.c
> > +index 3237fc8..2a38953 100644
> > +--- a/drivers/infiniband/hw/cxgb3/iwch_cm.c
> > ++++ b/drivers/infiniband/hw/cxgb3/iwch_cm.c
> > +@@ -234,7 +234,7 @@ static void *alloc_ep(int size, gfp_t gf
> > + epc = kmalloc(size, gfp);
> > + if (epc) {
> > + memset(epc, 0, size);
> > +- kref_init(&epc->kref);
> > ++ kref_init(&epc->kref, __free_ep);
> > + spin_lock_init(&epc->lock);
> > + init_waitqueue_head(&epc->waitq);
> > + }
> > +@@ -338,7 +338,7 @@ static struct rtable *find_route(struct
> > + }
> > + };
> > +
> > +- if (ip_route_output_flow(&rt, &fl, NULL, 0))
> > ++ if (ip_route_output_key(&rt, &fl))
> > + return NULL;
> > + return rt;
> > + }
> > +diff --git a/drivers/infiniband/hw/cxgb3/iwch_cm.h
> > b/drivers/infiniband/hw/cxgb3/iwch_cm.h
> > +index 893f9d0..e54e202 100644
> > +--- a/drivers/infiniband/hw/cxgb3/iwch_cm.h
> > ++++ b/drivers/infiniband/hw/cxgb3/iwch_cm.h
> > +@@ -57,7 +57,7 @@ #define MPA_FLAGS_MASK 0xE0
> > + #define put_ep(ep) { \
> > + PDBG("put_ep (via %s:%u) ep %p refcnt %d\n", __FUNCTION__, __LINE__, \
> > + ep, atomic_read(&((ep)->kref.refcount))); \
> > +- kref_put(&((ep)->kref), __free_ep); \
> > ++ kref_put(&((ep)->kref)); \
> > + }
> > +
> > + #define get_ep(ep) { \
>
> It's not easy to do something abou t kref_put (we'd need to define our own
> struct
> for it, like we did for work_struct).
> But surely ip_route_output_key can just be a macro in kernel_addons?
>
It could, but I was worried that if we did that and someone used it
expecting the sock and flags parameters to mean something, then they'd
get hosed.
And since the kref stuff was already being patched, I just went ahead
and used ip_route_output_key directly.
But I'll go ahead and add a ip_route_output_flow() backport.
_______________________________________________
openib-general mailing list
[email protected]
http://openib.org/mailman/listinfo/openib-general
To unsubscribe, please visit http://openib.org/mailman/listinfo/openib-general
