* [TRIVIAL] [patch, 2.5] scsi_qla1280.c free on error path
@ 2003-01-06 3:53 Rusty Trivial Russell
0 siblings, 0 replies; 2+ messages in thread
From: Rusty Trivial Russell @ 2003-01-06 3:53 UTC (permalink / raw)
To: jes, linux-scsi
[ Jes, is this OK? ]
From: Marcus Alanen <maalanen@ra.abo.fi>
Remove check_region in favour of request_region. Free resources
properly on error path. Horribly subtle ioremap/iounmap lurks here I
think, in qla1280_pci_config(), which the below patch should take care
of.
I'm wondering if there couldn't / shouldn't be a better way to
allocate resources. Obviously lots of drivers have broken error paths.
Is this even necessary?
Marcus
#
# create_patch: qla1280_release_on_error_path-2002-12-08-A.patch
# Date: Sun Dec 8 22:32:33 EET 2002
#
--- trivial-2.5-bk/drivers/scsi/qla1280.c.orig 2003-01-06 14:10:52.000000000 +1100
+++ trivial-2.5-bk/drivers/scsi/qla1280.c 2003-01-06 14:10:52.000000000 +1100
@@ -866,19 +866,17 @@
"qla1280", ha)) {
printk("qla1280 : Failed to reserve interrupt %d already "
"in use\n", host->irq);
- goto error_mem_alloced;
+ goto error_unmap;
}
#if !MEMORY_MAPPED_IO
/* Register the I/O space with Linux */
- if (check_region(host->io_port, 0xff)) {
+ if (!request_region(host->io_port, 0xff, "qla1280")) {
printk("qla1280 : Failed to reserve i/o region 0x%04lx-0x%04lx"
" already in use\n",
host->io_port, host->io_port + 0xff);
- free_irq(host->irq, ha);
- goto error_mem_alloced;
+ goto error_irq;
}
- request_region(host->io_port, 0xff, "qla1280");
#endif
reg = ha->iobase;
@@ -886,7 +884,7 @@
/* load the F/W, read paramaters, and init the H/W */
if (qla1280_initialize_adapter(ha)) {
printk(KERN_INFO "qla1x160:Failed to initialize adapter\n");
- goto error_mem_alloced;
+ goto error_region;
}
/* set our host ID (need to do something about our two IDs) */
@@ -894,6 +892,21 @@
return host;
+ error_region:
+#if !MEMORY_MAPPED_IO
+ release_region(host->io_port, 0xff);
+#endif
+
+ error_irq:
+ free_irq(host->irq, ha);
+
+ error_unmap:
+#if MEMORY_MAPPED_IO
+ if (ha->mmpbase)
+ iounmap((void *)(((unsigned long) ha->mmpbase) & PAGE_MASK));
+#endif
+
+
error_mem_alloced:
qla1280_mem_free(ha);
--
Don't blame me: the Monkey is driving
File: Marcus Alanen <maalanen@ra.abo.fi>: [patch, 2.5] scsi_qla1280.c free on error path
^ permalink raw reply [flat|nested] 2+ messages in thread* [TRIVIAL] [patch, 2.5] scsi_qla1280.c free on error path
@ 2003-02-06 5:48 Rusty Trivial Russell
0 siblings, 0 replies; 2+ messages in thread
From: Rusty Trivial Russell @ 2003-02-06 5:48 UTC (permalink / raw)
To: Linus Torvalds; +Cc: Jes Sorensen, linux-scsi
From: Marcus Alanen <maalanen@ra.abo.fi>
Remove check_region in favour of request_region. Free resources
properly on error path. Horribly subtle ioremap/iounmap lurks here I
think, in qla1280_pci_config(), which the below patch should take care
of.
I'm wondering if there couldn't / shouldn't be a better way to
allocate resources. Obviously lots of drivers have broken error paths.
Is this even necessary?
Marcus
#
# create_patch: qla1280_release_on_error_path-2002-12-08-A.patch
# Date: Sun Dec 8 22:32:33 EET 2002
#
--- trivial-2.5-bk/drivers/scsi/qla1280.c.orig 2003-02-06 16:30:17.000000000 +1100
+++ trivial-2.5-bk/drivers/scsi/qla1280.c 2003-02-06 16:30:17.000000000 +1100
@@ -866,19 +866,17 @@
"qla1280", ha)) {
printk("qla1280 : Failed to reserve interrupt %d already "
"in use\n", host->irq);
- goto error_mem_alloced;
+ goto error_unmap;
}
#if !MEMORY_MAPPED_IO
/* Register the I/O space with Linux */
- if (check_region(host->io_port, 0xff)) {
+ if (!request_region(host->io_port, 0xff, "qla1280")) {
printk("qla1280 : Failed to reserve i/o region 0x%04lx-0x%04lx"
" already in use\n",
host->io_port, host->io_port + 0xff);
- free_irq(host->irq, ha);
- goto error_mem_alloced;
+ goto error_irq;
}
- request_region(host->io_port, 0xff, "qla1280");
#endif
reg = ha->iobase;
@@ -886,7 +884,7 @@
/* load the F/W, read paramaters, and init the H/W */
if (qla1280_initialize_adapter(ha)) {
printk(KERN_INFO "qla1x160:Failed to initialize adapter\n");
- goto error_mem_alloced;
+ goto error_region;
}
/* set our host ID (need to do something about our two IDs) */
@@ -894,6 +892,21 @@
return host;
+ error_region:
+#if !MEMORY_MAPPED_IO
+ release_region(host->io_port, 0xff);
+#endif
+
+ error_irq:
+ free_irq(host->irq, ha);
+
+ error_unmap:
+#if MEMORY_MAPPED_IO
+ if (ha->mmpbase)
+ iounmap((void *)(((unsigned long) ha->mmpbase) & PAGE_MASK));
+#endif
+
+
error_mem_alloced:
qla1280_mem_free(ha);
--
Don't blame me: the Monkey is driving
File: Marcus Alanen <maalanen@ra.abo.fi>: [patch, 2.5] scsi_qla1280.c free on error path
^ permalink raw reply [flat|nested] 2+ messages in thread
end of thread, other threads:[~2003-02-06 5:48 UTC | newest]
Thread overview: 2+ messages (download: mbox.gz follow: Atom feed
-- links below jump to the message on this page --
2003-01-06 3:53 [TRIVIAL] [patch, 2.5] scsi_qla1280.c free on error path Rusty Trivial Russell
-- strict thread matches above, loose matches on Subject: below --
2003-02-06 5:48 Rusty Trivial Russell
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox