s390: use kthread_run in qeth and lcs

PATCH NAK Frank Pavlic 2 participants 3 messages Apr 30, 2007

Subsystems (4)

Networking Network drivers s390 s390 network drivers

lore

Share & tools

lore

Frank Pavlic

author

2007-04-30 15:42 UTC #0 lore raw

PATCH 3/5 s390: use kthread_run in qeth and lcs

From: Eric Biederman <[email protected]>

This patch modifies the qeth_recover thread to be started with kthread_run not a combination of kernel_thread and daemonize. Resulting in slightly simpler and more maintainable code.

Signed-off-by Eric W. Biederman <[email protected]>

Signed-off-by Frank Pavlic <[email protected]>

---

 lcs.c       |    8 +++-----
 qeth_main.c |    4 ++--
 2 files changed, 5 insertions(+), 7 deletions(-)

drivers/s390/net/lcs.c +3 -5

index 08a994f..0300d87 100644
@@ -36,6 +36,7 @@
 #include <linux/in.h>
 #include <linux/igmp.h>
 #include <linux/delay.h>
+#include <linux/kthread.h>
 #include <net/arp.h>
 #include <net/ip.h>

@@ -1248,7 +1249,6 @@ lcs_register_mc_addresses(void *data)
 	struct in_device *in4_dev;
 
 	card = (struct lcs_card *) data;
-	daemonize("regipm");
 
 	if (!lcs_do_run_thread(card, LCS_SET_MC_THREAD))
 		return 0;

@@ -1728,11 +1728,10 @@ lcs_start_kernel_thread(struct work_struct *work)
 	struct lcs_card *card = container_of(work, struct lcs_card, kernel_thread_starter);
 	LCS_DBF_TEXT(5, trace, "krnthrd");
 	if (lcs_do_start_thread(card, LCS_RECOVERY_THREAD))
-		kernel_thread(lcs_recovery, (void *) card, SIGCHLD);
+		kthread_run(lcs_recovery, card, "lcs_recover");
 #ifdef CONFIG_IP_MULTICAST
 	if (lcs_do_start_thread(card, LCS_SET_MC_THREAD))
-		kernel_thread(lcs_register_mc_addresses,
-				(void *) card, SIGCHLD);
+		kernel_run(lcs_register_mc_addresses, card, "regipm");
 #endif
 }

@@ -2232,7 +2231,6 @@ lcs_recovery(void *ptr)
         int rc;
 
 	card = (struct lcs_card *) ptr;
-	daemonize("lcs_recover");
 
 	LCS_DBF_TEXT(4, trace, "recover1");
 	if (!lcs_do_run_thread(card, LCS_RECOVERY_THREAD))

drivers/s390/net/qeth_main.c +2 -2

index e0eb113..0d10ee0 100644
@@ -50,6 +50,7 @@
 #include <linux/mii.h>
 #include <linux/rcupdate.h>
 #include <linux/ethtool.h>
+#include <linux/kthread.h>
 
 #include <net/arp.h>
 #include <net/ip.h>

@@ -957,7 +958,6 @@ qeth_recover(void *ptr)
 	int rc = 0;
 
 	card = (struct qeth_card *) ptr;
-	daemonize("qeth_recover");
 	QETH_DBF_TEXT(trace,2,"recover1");
 	QETH_DBF_HEX(trace, 2, &card, sizeof(void *));
 	if (!qeth_do_run_thread(card, QETH_RECOVER_THREAD))

@@ -1014,7 +1014,7 @@ qeth_start_kernel_thread(struct work_struct *work)
 	    card->write.state != CH_STATE_UP)
 		return;
 	if (qeth_do_start_thread(card, QETH_RECOVER_THREAD))
-		kernel_thread(qeth_recover, (void *) card, SIGCHLD);
+		kthread_run(qeth_recover, card, "qeth_recover");
 }

Signature · 1 line

1.5.1.2

Christoph Hellwig

NAK

2007-04-30 16:03 UTC #1 lore raw

Quoted 6 lines · Frank Pavlic

From: Eric Biederman <[email protected]>
This patch modifies the qeth_recover thread to be started with kthread_run not a combination of kernel_thread and daemonize. Resulting in slightly simpler and more maintainable code.

NACK, this is Eric's partial conversion which doesn't help much at all. In fact it might actually hurt by introducing unpaired kthread calls.

Frank Pavlic

author

2007-04-30 16:27 UTC #2 lore raw

Quoted 11 lines · Christoph Hellwig

Quoted 6 lines · Frank Pavlic
From: Eric Biederman <[email protected]>
This patch modifies the qeth_recover thread to be started with kthread_run not a combination of kernel_thread and daemonize. Resulting in slightly simpler and more maintainable code.
NACK, this is Eric's partial conversion which doesn't help much at all. In fact it might actually hurt by introducing unpaired kthread calls.

Alright Christoph, thanks for clarifying this..

@Jeff,
I'll send you another set without this one...

Frank