mirror of https://lore.kernel.org/lkml/
 help / color / mirror / Atom feed
* [PATCH][1/6] IB: Keep MAD work completion valid
@ 2005-04-01 18:23 Roland Dreier
  2005-04-01 18:23 ` [PATCH][2/6] IB: remove unneeded includes Roland Dreier
  0 siblings, 1 reply; 6+ messages in thread
From: Roland Dreier @ 2005-04-01 18:23 UTC (permalink / raw)
  To: akpm; +Cc: linux-kernel, openib-general

From: Sean Hefty <sean.hefty@intel.com>

Replace the *wc field in ib_mad_recv_wc from pointing to a structure
on the stack to one allocated with the received MAD buffer.  This
allows a client to access the *wc field after their receive completion
handler has returned.

Signed-off-by: Sean Hefty <sean.hefty@intel.com>
Signed-off-by: Roland Dreier <roland@topspin.com>


--- linux-export.orig/drivers/infiniband/core/mad.c	2005-03-31 19:07:01.000000000 -0800
+++ linux-export/drivers/infiniband/core/mad.c	2005-04-01 10:08:54.939957801 -0800
@@ -1600,7 +1600,8 @@
 			 DMA_FROM_DEVICE);
 
 	/* Setup MAD receive work completion from "normal" work completion */
-	recv->header.recv_wc.wc = wc;
+	recv->header.wc = *wc;
+	recv->header.recv_wc.wc = &recv->header.wc;
 	recv->header.recv_wc.mad_len = sizeof(struct ib_mad);
 	recv->header.recv_wc.recv_buf.mad = &recv->mad.mad;
 	recv->header.recv_wc.recv_buf.grh = &recv->grh;
--- linux-export.orig/drivers/infiniband/core/mad_priv.h	2005-03-31 19:07:14.000000000 -0800
+++ linux-export/drivers/infiniband/core/mad_priv.h	2005-04-01 10:08:54.961953027 -0800
@@ -69,6 +69,7 @@
 struct ib_mad_private_header {
 	struct ib_mad_list_head mad_list;
 	struct ib_mad_recv_wc recv_wc;
+	struct ib_wc wc;
 	DECLARE_PCI_UNMAP_ADDR(mapping)
 } __attribute__ ((packed));
 


^ permalink raw reply	[flat|nested] 6+ messages in thread

* [PATCH][2/6] IB: remove unneeded includes
  2005-04-01 18:23 [PATCH][1/6] IB: Keep MAD work completion valid Roland Dreier
@ 2005-04-01 18:23 ` Roland Dreier
  2005-04-01 18:23   ` [PATCH][3/6] IB: Fix FMR pool crash Roland Dreier
  0 siblings, 1 reply; 6+ messages in thread
From: Roland Dreier @ 2005-04-01 18:23 UTC (permalink / raw)
  To: akpm; +Cc: linux-kernel, openib-general

From: Hal Rosenstock <halr@voltaire.com>

Eliminate no longer needed include files

Signed-off-by: Hal Rosenstock <halr@voltaire.com>
Signed-off-by: Roland Dreier <roland@topspin.com>


--- linux-export.orig/drivers/infiniband/core/mad.c	2005-04-01 10:08:54.939957801 -0800
+++ linux-export/drivers/infiniband/core/mad.c	2005-04-01 10:08:56.473624910 -0800
@@ -33,9 +33,6 @@
  */
 
 #include <linux/dma-mapping.h>
-#include <linux/interrupt.h>
-
-#include <ib_mad.h>
 
 #include "mad_priv.h"
 #include "smi.h"


^ permalink raw reply	[flat|nested] 6+ messages in thread

* [PATCH][3/6] IB: Fix FMR pool crash
  2005-04-01 18:23 ` [PATCH][2/6] IB: remove unneeded includes Roland Dreier
@ 2005-04-01 18:23   ` Roland Dreier
  2005-04-01 18:23     ` [PATCH][4/6] IB: Trivial FMR printk cleanup Roland Dreier
  0 siblings, 1 reply; 6+ messages in thread
From: Roland Dreier @ 2005-04-01 18:23 UTC (permalink / raw)
  To: akpm; +Cc: linux-kernel, openib-general

Mask bits correctly from jhash result in ib_fmr_hash() so that the
computed bucket index is within our hash table.  This fixes an SDP
crash.

Signed-off-by: Roland Dreier <roland@topspin.com>


--- linux-export.orig/drivers/infiniband/core/fmr_pool.c	2005-03-31 19:07:05.000000000 -0800
+++ linux-export/drivers/infiniband/core/fmr_pool.c	2005-04-01 10:08:58.240241456 -0800
@@ -103,9 +103,8 @@
 
 static inline u32 ib_fmr_hash(u64 first_page)
 {
-	return jhash_2words((u32) first_page,
-			    (u32) (first_page >> 32),
-			    0);
+	return jhash_2words((u32) first_page, (u32) (first_page >> 32), 0) &
+		(IB_FMR_HASH_SIZE - 1);
 }
 
 /* Caller must hold pool_lock */


^ permalink raw reply	[flat|nested] 6+ messages in thread

* [PATCH][4/6] IB: Trivial FMR printk cleanup
  2005-04-01 18:23   ` [PATCH][3/6] IB: Fix FMR pool crash Roland Dreier
@ 2005-04-01 18:23     ` Roland Dreier
  2005-04-01 18:23       ` [PATCH][5/6] IB: Fix user MAD registrations with class 0 Roland Dreier
  0 siblings, 1 reply; 6+ messages in thread
From: Roland Dreier @ 2005-04-01 18:23 UTC (permalink / raw)
  To: akpm; +Cc: linux-kernel, openib-general

From: Libor Michalek <libor@topspin.com>

Add missing newline in printk.

Signed-off-by: Libor Michalek <libor@topspin.com>
Signed-off-by: Roland Dreier <roland@topspin.com>


--- linux-export.orig/drivers/infiniband/core/fmr_pool.c	2005-04-01 10:08:58.240241456 -0800
+++ linux-export/drivers/infiniband/core/fmr_pool.c	2005-04-01 10:08:59.539959345 -0800
@@ -442,7 +442,7 @@
 		list_add(&fmr->list, &pool->free_list);
 		spin_unlock_irqrestore(&pool->pool_lock, flags);
 
-		printk(KERN_WARNING "fmr_map returns %d",
+		printk(KERN_WARNING "fmr_map returns %d\n",
 		       result);
 
 		return ERR_PTR(result);


^ permalink raw reply	[flat|nested] 6+ messages in thread

* [PATCH][5/6] IB: Fix user MAD registrations with class 0
  2005-04-01 18:23     ` [PATCH][4/6] IB: Trivial FMR printk cleanup Roland Dreier
@ 2005-04-01 18:23       ` Roland Dreier
  2005-04-01 18:23         ` [PATCH][6/6] IB: Remove incorrect comments Roland Dreier
  0 siblings, 1 reply; 6+ messages in thread
From: Roland Dreier @ 2005-04-01 18:23 UTC (permalink / raw)
  To: akpm; +Cc: linux-kernel, openib-general

Fix handling of MAD agent registrations with mgmt_class == 0.  In this
case ib_umad should pass a NULL registration request to the MAD core
rather than a request with mgmt_class set to 0.

Signed-off-by: Roland Dreier <roland@topspin.com>


--- linux-export.orig/drivers/infiniband/core/user_mad.c	2005-03-31 19:06:42.000000000 -0800
+++ linux-export/drivers/infiniband/core/user_mad.c	2005-04-01 10:09:01.250588043 -0800
@@ -389,15 +389,17 @@
 	goto out;
 
 found:
-	req.mgmt_class         = ureq.mgmt_class;
-	req.mgmt_class_version = ureq.mgmt_class_version;
-	memcpy(req.method_mask, ureq.method_mask, sizeof req.method_mask);
-	memcpy(req.oui,         ureq.oui,         sizeof req.oui);
+	if (ureq.mgmt_class) {
+		req.mgmt_class         = ureq.mgmt_class;
+		req.mgmt_class_version = ureq.mgmt_class_version;
+		memcpy(req.method_mask, ureq.method_mask, sizeof req.method_mask);
+		memcpy(req.oui,         ureq.oui,         sizeof req.oui);
+	}
 
 	agent = ib_register_mad_agent(file->port->ib_dev, file->port->port_num,
 				      ureq.qpn ? IB_QPT_GSI : IB_QPT_SMI,
-				      &req, 0, send_handler, recv_handler,
-				      file);
+				      ureq.mgmt_class ? &req : NULL,
+				      0, send_handler, recv_handler, file);
 	if (IS_ERR(agent)) {
 		ret = PTR_ERR(agent);
 		goto out;


^ permalink raw reply	[flat|nested] 6+ messages in thread

* [PATCH][6/6] IB: Remove incorrect comments
  2005-04-01 18:23       ` [PATCH][5/6] IB: Fix user MAD registrations with class 0 Roland Dreier
@ 2005-04-01 18:23         ` Roland Dreier
  0 siblings, 0 replies; 6+ messages in thread
From: Roland Dreier @ 2005-04-01 18:23 UTC (permalink / raw)
  To: akpm; +Cc: linux-kernel, openib-general

From: Hal Rosenstock <halr@voltaire.com>

Eliminate unneeded and misleading comments

Signed-off-by: Hal Rosenstock <halr@voltaire.com>
Signed-off-by: Roland Dreier <roland@topspin.com>


--- linux-export.orig/drivers/infiniband/core/agent.c	2005-03-31 19:06:48.000000000 -0800
+++ linux-export/drivers/infiniband/core/agent.c	2005-04-01 10:09:02.621290525 -0800
@@ -129,7 +129,6 @@
 		goto out;
 	agent_send_wr->mad = mad_priv;
 
-	/* PCI mapping */
 	gather_list.addr = dma_map_single(mad_agent->device->dma_device,
 					  &mad_priv->mad,
 					  sizeof(mad_priv->mad),
@@ -261,7 +260,6 @@
 	list_del(&agent_send_wr->send_list);
 	spin_unlock_irqrestore(&port_priv->send_list_lock, flags);
 
-	/* Unmap PCI */
 	dma_unmap_single(mad_agent->device->dma_device,
 			 pci_unmap_addr(agent_send_wr, mapping),
 			 sizeof(agent_send_wr->mad->mad),
--- linux-export.orig/drivers/infiniband/core/mad.c	2005-04-01 10:08:56.473624910 -0800
+++ linux-export/drivers/infiniband/core/mad.c	2005-04-01 10:09:02.768258624 -0800
@@ -2283,7 +2283,6 @@
 		/* Remove from posted receive MAD list */
 		list_del(&mad_list->list);
 
-		/* Undo PCI mapping */
 		dma_unmap_single(qp_info->port_priv->device->dma_device,
 				 pci_unmap_addr(&recv->header, mapping),
 				 sizeof(struct ib_mad_private) -


^ permalink raw reply	[flat|nested] 6+ messages in thread

end of thread, other threads:[~2005-04-01 18:51 UTC | newest]

Thread overview: 6+ messages (download: mbox.gz / follow: Atom feed)
-- links below jump to the message on this page --
2005-04-01 18:23 [PATCH][1/6] IB: Keep MAD work completion valid Roland Dreier
2005-04-01 18:23 ` [PATCH][2/6] IB: remove unneeded includes Roland Dreier
2005-04-01 18:23   ` [PATCH][3/6] IB: Fix FMR pool crash Roland Dreier
2005-04-01 18:23     ` [PATCH][4/6] IB: Trivial FMR printk cleanup Roland Dreier
2005-04-01 18:23       ` [PATCH][5/6] IB: Fix user MAD registrations with class 0 Roland Dreier
2005-04-01 18:23         ` [PATCH][6/6] IB: Remove incorrect comments Roland Dreier

This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox

all inboxes | Powered by JetHome®