From: Benjamin Herrenschmidt <benh@kernel.crashing.org>
To: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Thomas Meyer <thomas@m3y3r.de>, Ingo Molnar <mingo@elte.hu>,
Stefan Richter <stefanr@s5r6.in-berlin.de>,
Thomas Gleixner <tglx@linutronix.de>,
Ivan Kokshaysky <ink@jurassic.park.msu.ru>,
"Rafael J. Wysocki" <rjw@sisk.pl>,
LKML <linux-kernel@vger.kernel.org>,
Adrian Bunk <bunk@kernel.org>,
Andrew Morton <akpm@linux-foundation.org>,
Natalie Protasevich <protasnb@gmail.com>
Subject: Re: ohci1394 problem (MMIO broken) (was 2.6.25-rc6-git6: Reported regressions from 2.6.24)
Date: Thu, 27 Mar 2008 10:25:46 +1100 [thread overview]
Message-ID: <1206573946.6926.51.camel@pasglop> (raw)
In-Reply-To: <1206502750.10573.12.camel@pasglop>
While at it, here's an updated version of my patch that applies on top
of yesterday git. I also turned "len" into a resource_size_t as it's
technically the right type to use in the generic code, even if it makes
little sense to try to map more than 32 bits on a 32 bits arch :-) Maybe
we should actually add some kind of warning here or error checking
though rather than just cropping the value.
I "fixed" all archs even those that don't technically need it (like real
64 bits one) for consistency (and who knows if they ever want to make
resource_size_t even bigger ...). The patch actually fixes at least one
bug in the case of arch/ppc which needs that change.
---
iomap: Fix 64 bits resources on 32 bits
Almost all implementations of pci_iomap() in the kernel, including
the generic lib/iomap.c one, copies the content of a struct resource
into unsigned long's which will break on 32 bits platforms with 64
bits resources.
This fixes all definitions of pci_iomap() to use resource_size_t. I
also "fixed" the 64bits arch for consistency.
Signed-off-by: Benjamin Herrenschmidt <benh@kernel.crashing.org>
---
Index: linux-work/lib/iomap.c
===================================================================
--- linux-work.orig/lib/iomap.c 2008-03-26 10:39:25.000000000 +1100
+++ linux-work/lib/iomap.c 2008-03-27 10:20:14.000000000 +1100
@@ -257,7 +257,7 @@ EXPORT_SYMBOL(ioport_unmap);
void __iomem *pci_iomap(struct pci_dev *dev, int bar, unsigned long maxlen)
{
resource_size_t start = pci_resource_start(dev, bar);
- unsigned long len = pci_resource_len(dev, bar);
+ resource_size_t len = pci_resource_len(dev, bar);
unsigned long flags = pci_resource_flags(dev, bar);
if (!len || !start)
Index: linux-work/arch/alpha/kernel/pci.c
===================================================================
--- linux-work.orig/arch/alpha/kernel/pci.c 2007-09-28 11:42:04.000000000 +1000
+++ linux-work/arch/alpha/kernel/pci.c 2008-03-27 10:19:59.000000000 +1100
@@ -547,8 +547,8 @@ sys_pciconfig_iobase(long which, unsigne
void __iomem *pci_iomap(struct pci_dev *dev, int bar, unsigned long maxlen)
{
- unsigned long start = pci_resource_start(dev, bar);
- unsigned long len = pci_resource_len(dev, bar);
+ resource_size_t start = pci_resource_start(dev, bar);
+ resource_size_t len = pci_resource_len(dev, bar);
unsigned long flags = pci_resource_flags(dev, bar);
if (!len || !start)
Index: linux-work/arch/arm/mm/iomap.c
===================================================================
--- linux-work.orig/arch/arm/mm/iomap.c 2007-07-27 13:44:41.000000000 +1000
+++ linux-work/arch/arm/mm/iomap.c 2008-03-27 10:19:59.000000000 +1100
@@ -26,8 +26,8 @@ EXPORT_SYMBOL(ioport_unmap);
#ifdef CONFIG_PCI
void __iomem *pci_iomap(struct pci_dev *dev, int bar, unsigned long maxlen)
{
- unsigned long start = pci_resource_start(dev, bar);
- unsigned long len = pci_resource_len(dev, bar);
+ resource_size_t start = pci_resource_start(dev, bar);
+ resource_size_t len = pci_resource_len(dev, bar);
unsigned long flags = pci_resource_flags(dev, bar);
if (!len || !start)
Index: linux-work/arch/frv/mb93090-mb00/pci-iomap.c
===================================================================
--- linux-work.orig/arch/frv/mb93090-mb00/pci-iomap.c 2007-07-27 13:44:41.000000000 +1000
+++ linux-work/arch/frv/mb93090-mb00/pci-iomap.c 2008-03-27 10:19:59.000000000 +1100
@@ -13,8 +13,8 @@
void __iomem *pci_iomap(struct pci_dev *dev, int bar, unsigned long maxlen)
{
- unsigned long start = pci_resource_start(dev, bar);
- unsigned long len = pci_resource_len(dev, bar);
+ resource_size_t start = pci_resource_start(dev, bar);
+ resource_size_t len = pci_resource_len(dev, bar);
unsigned long flags = pci_resource_flags(dev, bar);
if (!len || !start)
Index: linux-work/arch/mips/lib/iomap-pci.c
===================================================================
--- linux-work.orig/arch/mips/lib/iomap-pci.c 2007-09-28 11:42:05.000000000 +1000
+++ linux-work/arch/mips/lib/iomap-pci.c 2008-03-27 10:19:59.000000000 +1100
@@ -45,8 +45,8 @@ static void __iomem *ioport_map_pci(stru
*/
void __iomem *pci_iomap(struct pci_dev *dev, int bar, unsigned long maxlen)
{
- unsigned long start = pci_resource_start(dev, bar);
- unsigned long len = pci_resource_len(dev, bar);
+ resource_size_t start = pci_resource_start(dev, bar);
+ resource_size_t len = pci_resource_len(dev, bar);
unsigned long flags = pci_resource_flags(dev, bar);
if (!len || !start)
Index: linux-work/arch/parisc/lib/iomap.c
===================================================================
--- linux-work.orig/arch/parisc/lib/iomap.c 2007-07-27 13:44:42.000000000 +1000
+++ linux-work/arch/parisc/lib/iomap.c 2008-03-27 10:19:59.000000000 +1100
@@ -438,8 +438,8 @@ void ioport_unmap(void __iomem *addr)
/* Create a virtual mapping cookie for a PCI BAR (memory or IO) */
void __iomem *pci_iomap(struct pci_dev *dev, int bar, unsigned long maxlen)
{
- unsigned long start = pci_resource_start(dev, bar);
- unsigned long len = pci_resource_len(dev, bar);
+ resource_size_t start = pci_resource_start(dev, bar);
+ resource_size_t len = pci_resource_len(dev, bar);
unsigned long flags = pci_resource_flags(dev, bar);
if (!len || !start)
Index: linux-work/arch/sh/drivers/pci/pci.c
===================================================================
--- linux-work.orig/arch/sh/drivers/pci/pci.c 2008-03-19 13:52:27.000000000 +1100
+++ linux-work/arch/sh/drivers/pci/pci.c 2008-03-27 10:19:59.000000000 +1100
@@ -191,8 +191,8 @@ void __init pcibios_update_irq(struct pc
void __iomem *pci_iomap(struct pci_dev *dev, int bar, unsigned long maxlen)
{
- unsigned long start = pci_resource_start(dev, bar);
- unsigned long len = pci_resource_len(dev, bar);
+ resource_size_t start = pci_resource_start(dev, bar);
+ resource_size_t len = pci_resource_len(dev, bar);
unsigned long flags = pci_resource_flags(dev, bar);
if (unlikely(!len || !start))
Index: linux-work/arch/mn10300/unit-asb2305/pci-iomap.c
===================================================================
--- linux-work.orig/arch/mn10300/unit-asb2305/pci-iomap.c 2008-03-19 13:52:26.000000000 +1100
+++ linux-work/arch/mn10300/unit-asb2305/pci-iomap.c 2008-03-27 10:19:59.000000000 +1100
@@ -16,8 +16,8 @@
*/
void __iomem *pci_iomap(struct pci_dev *dev, int bar, unsigned long maxlen)
{
- unsigned long start = pci_resource_start(dev, bar);
- unsigned long len = pci_resource_len(dev, bar);
+ resource_size_t start = pci_resource_start(dev, bar);
+ resource_size_t len = pci_resource_len(dev, bar);
unsigned long flags = pci_resource_flags(dev, bar);
if (!len || !start)
Index: linux-work/arch/ppc/kernel/pci.c
===================================================================
--- linux-work.orig/arch/ppc/kernel/pci.c 2007-09-28 11:42:05.000000000 +1000
+++ linux-work/arch/ppc/kernel/pci.c 2008-03-27 10:19:59.000000000 +1100
@@ -1176,8 +1176,8 @@ void __init pci_init_resource(struct res
void __iomem *pci_iomap(struct pci_dev *dev, int bar, unsigned long max)
{
- unsigned long start = pci_resource_start(dev, bar);
- unsigned long len = pci_resource_len(dev, bar);
+ resource_size_t start = pci_resource_start(dev, bar);
+ resource_size_t len = pci_resource_len(dev, bar);
unsigned long flags = pci_resource_flags(dev, bar);
if (!len)
Index: linux-work/arch/sparc/lib/iomap.c
===================================================================
--- linux-work.orig/arch/sparc/lib/iomap.c 2007-07-27 13:44:42.000000000 +1000
+++ linux-work/arch/sparc/lib/iomap.c 2008-03-27 10:19:59.000000000 +1100
@@ -21,8 +21,8 @@ EXPORT_SYMBOL(ioport_unmap);
/* Create a virtual mapping cookie for a PCI BAR (memory or IO) */
void __iomem *pci_iomap(struct pci_dev *dev, int bar, unsigned long maxlen)
{
- unsigned long start = pci_resource_start(dev, bar);
- unsigned long len = pci_resource_len(dev, bar);
+ resource_size_t start = pci_resource_start(dev, bar);
+ resource_size_t len = pci_resource_len(dev, bar);
unsigned long flags = pci_resource_flags(dev, bar);
if (!len || !start)
Index: linux-work/arch/sparc64/lib/iomap.c
===================================================================
--- linux-work.orig/arch/sparc64/lib/iomap.c 2007-07-27 13:44:42.000000000 +1000
+++ linux-work/arch/sparc64/lib/iomap.c 2008-03-27 10:19:59.000000000 +1100
@@ -21,8 +21,8 @@ EXPORT_SYMBOL(ioport_unmap);
/* Create a virtual mapping cookie for a PCI BAR (memory or IO) */
void __iomem *pci_iomap(struct pci_dev *dev, int bar, unsigned long maxlen)
{
- unsigned long start = pci_resource_start(dev, bar);
- unsigned long len = pci_resource_len(dev, bar);
+ resource_size_t start = pci_resource_start(dev, bar);
+ resource_size_t len = pci_resource_len(dev, bar);
unsigned long flags = pci_resource_flags(dev, bar);
if (!len || !start)
Index: linux-work/arch/v850/kernel/rte_mb_a_pci.c
===================================================================
--- linux-work.orig/arch/v850/kernel/rte_mb_a_pci.c 2008-03-03 11:58:41.000000000 +1100
+++ linux-work/arch/v850/kernel/rte_mb_a_pci.c 2008-03-27 10:19:59.000000000 +1100
@@ -790,8 +790,8 @@ pci_free_consistent (struct pci_dev *pde
void __iomem *pci_iomap (struct pci_dev *dev, int bar, unsigned long max)
{
- unsigned long start = pci_resource_start (dev, bar);
- unsigned long len = pci_resource_len (dev, bar);
+ resource_size_t start = pci_resource_start (dev, bar);
+ resource_size_t len = pci_resource_len (dev, bar);
if (!start || len == 0)
return 0;
next prev parent reply other threads:[~2008-03-26 23:27 UTC|newest]
Thread overview: 97+ messages / expand[flat|nested] mbox.gz Atom feed top
2008-03-22 1:59 2.6.25-rc6-git6: Reported regressions from 2.6.24 Rafael J. Wysocki
2008-03-22 2:52 ` Jeff Garzik
2008-03-22 11:11 ` Ingo Molnar
2008-03-22 5:33 ` Andrew Morton
2008-03-22 11:15 ` Ingo Molnar
2008-03-22 17:53 ` Rafael J. Wysocki
2008-03-22 16:32 ` Heiko Carstens
2008-03-22 17:56 ` Rafael J. Wysocki
2008-03-22 16:34 ` ohci1394 problem (MMIO broken) (was 2.6.25-rc6-git6: Reported regressions from 2.6.24) Stefan Richter
2008-03-22 17:24 ` Thomas Meyer
2008-03-22 17:58 ` Rafael J. Wysocki
2008-03-22 18:27 ` Stefan Richter
2008-03-22 19:02 ` Stefan Richter
2008-03-22 21:33 ` Linus Torvalds
2008-03-22 21:58 ` Thomas Gleixner
2008-03-22 21:59 ` H. Peter Anvin
2008-03-22 22:27 ` ohci1394 problem (MMIO broken) (was 2.6.25-rc6-git6: Reported regressions from 2.6.24) [Bug 10080] Stefan Richter
2008-03-22 23:28 ` ohci1394 problem (MMIO broken) (was 2.6.25-rc6-git6: Reported regressions from 2.6.24) Yinghai Lu
2008-03-22 23:31 ` H. Peter Anvin
2008-03-23 2:00 ` Yinghai Lu
2008-03-23 2:39 ` H. Peter Anvin
2008-03-23 3:30 ` Yinghai Lu
2008-03-22 23:01 ` Linus Torvalds
2008-03-22 23:50 ` Stefan Richter
2008-03-23 6:35 ` Ingo Molnar
2008-03-24 19:34 ` Thomas Meyer
2008-03-24 19:47 ` Ingo Molnar
2008-03-24 20:17 ` Thomas Meyer
2008-03-24 20:47 ` Linus Torvalds
2008-03-26 3:39 ` Benjamin Herrenschmidt
2008-03-26 23:25 ` Benjamin Herrenschmidt [this message]
2008-03-24 19:58 ` Thomas Meyer
2008-03-24 20:50 ` Linus Torvalds
2008-03-24 21:24 ` Stefan Richter
2008-03-24 21:47 ` Stefan Richter
2008-03-25 7:31 ` Ingo Molnar
2008-03-25 16:50 ` Ingo Molnar
2008-03-25 17:06 ` Stefan Richter
2008-03-25 18:32 ` Thomas Meyer
2008-03-25 20:11 ` Ingo Molnar
2008-03-25 20:29 ` Ingo Molnar
2008-03-25 21:08 ` Thomas Meyer
2008-03-25 21:47 ` Linus Torvalds
2008-03-25 22:02 ` Thomas Meyer
2008-03-26 10:14 ` Ivan Kokshaysky
2008-03-26 12:17 ` Benjamin Herrenschmidt
2008-03-26 13:54 ` [patch] pci: revert "PCI: remove transparent bridge sizing" Ingo Molnar
2008-03-26 17:45 ` Thomas Meyer
2008-03-26 18:07 ` Gary Hade
2008-03-26 18:33 ` Linus Torvalds
2008-03-26 20:30 ` Gary Hade
2008-03-26 20:46 ` Linus Torvalds
2008-03-26 20:58 ` Ivan Kokshaysky
2008-03-26 21:41 ` Linus Torvalds
2008-03-26 21:57 ` Benjamin Herrenschmidt
2008-03-26 22:07 ` Alan Cox
2008-03-26 22:27 ` Benjamin Herrenschmidt
2008-03-26 22:10 ` Ingo Molnar
2008-03-26 22:29 ` Benjamin Herrenschmidt
2008-03-26 22:47 ` Linus Torvalds
2008-03-26 22:54 ` Benjamin Herrenschmidt
2008-03-26 23:18 ` Benjamin Herrenschmidt
2008-03-26 22:30 ` Ingo Molnar
2008-03-26 22:31 ` Linus Torvalds
2008-03-26 22:47 ` Alan Cox
2008-03-26 23:17 ` Benjamin Herrenschmidt
2008-03-26 23:29 ` Ivan Kokshaysky
2008-03-26 23:43 ` Linus Torvalds
2008-03-27 17:12 ` Linus Torvalds
2008-03-27 22:18 ` Ivan Kokshaysky
2008-03-27 22:34 ` Linus Torvalds
2008-03-28 19:24 ` Gary Hade
2008-03-28 20:46 ` Gary Hade
2008-03-30 15:44 ` Ivan Kokshaysky
2008-03-26 23:45 ` Benjamin Herrenschmidt
2008-03-26 21:29 ` Ingo Molnar
2008-03-26 11:12 ` ohci1394 problem (MMIO broken) (was 2.6.25-rc6-git6: Reported regressions from 2.6.24) Ivan Kokshaysky
2008-03-25 21:02 ` Thomas Meyer
2008-03-25 23:33 ` Benjamin Herrenschmidt
2008-03-26 0:03 ` Yinghai Lu
2008-03-26 0:12 ` Benjamin Herrenschmidt
2008-03-23 12:57 ` 2.6.25-rc6-git6: Reported regressions from 2.6.24 Alan Cox
2008-03-26 16:30 ` Ray Lee
2008-03-26 17:02 ` Adrian Bunk
2008-03-26 22:11 ` Rafael J. Wysocki
2008-03-27 10:18 ` Romano Giannetti
2008-03-27 14:43 ` Ray Lee
2008-03-31 18:21 ` Linus Torvalds
2008-03-31 19:29 ` Mark Lord
2008-03-31 21:04 ` Tino Keitel
2008-03-31 21:26 ` Tino Keitel
2008-04-03 19:06 ` 2.6.25-rc7/8: Another resume regression Mark Lord
2008-04-05 2:27 ` Mark Lord
2008-04-07 10:51 ` Rafael J. Wysocki
2008-04-07 15:51 ` Mark Lord
2008-04-07 17:40 ` Rafael J. Wysocki
2008-04-08 15:35 ` Mark Lord
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=1206573946.6926.51.camel@pasglop \
--to=benh@kernel.crashing.org \
--cc=akpm@linux-foundation.org \
--cc=bunk@kernel.org \
--cc=ink@jurassic.park.msu.ru \
--cc=linux-kernel@vger.kernel.org \
--cc=mingo@elte.hu \
--cc=protasnb@gmail.com \
--cc=rjw@sisk.pl \
--cc=stefanr@s5r6.in-berlin.de \
--cc=tglx@linutronix.de \
--cc=thomas@m3y3r.de \
--cc=torvalds@linux-foundation.org \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).