From: ebiederm@xmission.com (Eric W. Biederman)
To: Andrew Morton <akpm@linux-foundation.org>,
Linus Torvalds <torvalds@linux-foundation.org>
Cc: "Kok, Auke" <auke-jan.h.kok@intel.com>,
Michal Piotrowski <michal.k.k.piotrowski@gmail.com>,
Jeff Garzik <jeff@garzik.org>,
Greg Kroah-Hartman <gregkh@suse.de>,
linux-pm@lists.osdl.org,
Linux Kernel Mailing List <linux-kernel@vger.kernel.org>,
Adrian Bunk <bunk@stusta.de>,
michael@ellerman.id.au, linux-pci@atrey.karlin.mff.cuni.cz,
Pavel Machek <pavel@ucw.cz>, Jens Axboe <jens.axboe@oracle.com>,
"Michael S. Tsirkin" <mst@mellanox.co.il>,
Thomas Gleixner <tglx@linutronix.de>, Ingo Molnar <mingo@elte.hu>
Subject: [PATCH 2/2] pci: Repair pci_save/restore_state so we can restore one save many times.
Date: Thu, 08 Mar 2007 13:06:13 -0700 [thread overview]
Message-ID: <m1lki75wlm.fsf_-_@ebiederm.dsl.xmission.com> (raw)
In-Reply-To: <m1ps7j5wnq.fsf_-_@ebiederm.dsl.xmission.com> (Eric W. Biederman's message of "Thu, 08 Mar 2007 13:04:57 -0700")
Because we do not reserve space for the pci-x and pci-e state in struct
pci dev we need to dynamically allocate it. However because we need
to support restore being called multiple times after a single save
it is never safe to free the buffers we have allocated to hold the
state.
So this patch modifies the save routines to first check to see
if we have already allocated a state buffer before allocating
a new one. Then the restore routines are modified to not free
the state after restoring it. Simple and it fixes some subtle
error path handling bugs, that are hard to test for.
Signed-off-by: Eric W. Biederman <ebiederm@xmission.com>
---
drivers/pci/pci.c | 12 ++++++------
include/linux/pci.h | 5 -----
2 files changed, 6 insertions(+), 11 deletions(-)
diff --git a/drivers/pci/pci.c b/drivers/pci/pci.c
index 6fb78df..b292c9a 100644
--- a/drivers/pci/pci.c
+++ b/drivers/pci/pci.c
@@ -551,7 +551,9 @@ static int pci_save_pcie_state(struct pci_dev *dev)
if (pos <= 0)
return 0;
- save_state = kzalloc(sizeof(*save_state) + sizeof(u16) * 4, GFP_KERNEL);
+ save_state = pci_find_saved_cap(dev, PCI_CAP_ID_EXP);
+ if (!save_state)
+ save_state = kzalloc(sizeof(*save_state) + sizeof(u16) * 4, GFP_KERNEL);
if (!save_state) {
dev_err(&dev->dev, "Out of memory in pci_save_pcie_state\n");
return -ENOMEM;
@@ -582,8 +584,6 @@ static void pci_restore_pcie_state(struct pci_dev *dev)
pci_write_config_word(dev, pos + PCI_EXP_LNKCTL, cap[i++]);
pci_write_config_word(dev, pos + PCI_EXP_SLTCTL, cap[i++]);
pci_write_config_word(dev, pos + PCI_EXP_RTCTL, cap[i++]);
- pci_remove_saved_cap(save_state);
- kfree(save_state);
}
@@ -597,7 +597,9 @@ static int pci_save_pcix_state(struct pci_dev *dev)
if (pos <= 0)
return 0;
- save_state = kzalloc(sizeof(*save_state) + sizeof(u16), GFP_KERNEL);
+ save_state = pci_find_saved_cap(dev, PCI_CAP_ID_EXP);
+ if (!save_state)
+ save_state = kzalloc(sizeof(*save_state) + sizeof(u16), GFP_KERNEL);
if (!save_state) {
dev_err(&dev->dev, "Out of memory in pci_save_pcie_state\n");
return -ENOMEM;
@@ -622,8 +624,6 @@ static void pci_restore_pcix_state(struct pci_dev *dev)
cap = (u16 *)&save_state->data[0];
pci_write_config_word(dev, pos + PCI_X_CMD, cap[i++]);
- pci_remove_saved_cap(save_state);
- kfree(save_state);
}
diff --git a/include/linux/pci.h b/include/linux/pci.h
index 78417e4..481ea06 100644
--- a/include/linux/pci.h
+++ b/include/linux/pci.h
@@ -209,11 +209,6 @@ static inline void pci_add_saved_cap(struct pci_dev *pci_dev,
hlist_add_head(&new_cap->next, &pci_dev->saved_cap_space);
}
-static inline void pci_remove_saved_cap(struct pci_cap_saved_state *cap)
-{
- hlist_del(&cap->next);
-}
-
/*
* For PCI devices, the region numbers are assigned this way:
*
--
1.5.0.g53756
next prev parent reply other threads:[~2007-03-08 20:06 UTC|newest]
Thread overview: 104+ messages / expand[flat|nested] mbox.gz Atom feed top
[not found] <Pine.LNX.4.64.0702202043280.4043@woody.linux-foundation.org>
2007-02-25 17:52 ` 2.6.21-rc1: known regressions (part 1) Adrian Bunk
2007-02-28 18:16 ` Karasyov, Konstantin A
2007-02-25 17:55 ` 2.6.21-rc1: known regressions (part 2) Adrian Bunk
2007-02-27 10:02 ` Jens Axboe
2007-02-27 10:21 ` Pavel Machek
2007-02-27 10:30 ` Jens Axboe
2007-02-27 10:34 ` Ingo Molnar
2007-02-27 10:59 ` Jens Axboe
2007-02-27 11:15 ` Jens Axboe
2007-02-27 13:09 ` Jens Axboe
2007-03-01 9:34 ` Ingo Molnar
2007-03-01 10:41 ` Ingo Molnar
2007-03-01 14:52 ` Ingo Molnar
2007-03-01 16:12 ` Rafael J. Wysocki
2007-03-02 0:26 ` Linus Torvalds
2007-03-02 0:41 ` Linus Torvalds
2007-03-02 7:14 ` Ingo Molnar
2007-03-02 7:21 ` Ingo Molnar
2007-03-02 8:04 ` Ingo Molnar
2007-03-02 10:20 ` Ingo Molnar
2007-03-02 10:22 ` [patch] KVM: T60 resume fix Ingo Molnar
2007-03-02 11:39 ` Michael S. Tsirkin
2007-03-03 8:22 ` Avi Kivity
2007-03-03 8:21 ` Avi Kivity
2007-03-03 11:57 ` Andrew Morton
2007-03-03 12:07 ` Junio C Hamano
2007-03-05 8:22 ` Ingo Molnar
2007-03-05 8:50 ` Avi Kivity
2007-03-05 8:44 ` Ingo Molnar
2007-03-05 8:57 ` Ingo Molnar
2007-03-05 9:27 ` Avi Kivity
2007-03-05 10:05 ` Ingo Molnar
2007-03-05 10:33 ` Avi Kivity
2007-03-05 10:33 ` Ingo Molnar
2007-03-05 10:40 ` Michael S. Tsirkin
2007-03-05 12:54 ` Michael S. Tsirkin
2007-03-05 12:50 ` Ingo Molnar
2007-03-05 13:26 ` Michael S. Tsirkin
2007-03-05 13:32 ` Ingo Molnar
2007-03-05 10:23 ` Michael S. Tsirkin
2007-03-05 10:29 ` Ingo Molnar
2007-03-05 15:44 ` 2.6.21-rc1: known regressions (part 2) Michael S. Tsirkin
2007-03-05 16:14 ` Michael S. Tsirkin
2007-03-05 16:41 ` Ingo Molnar
2007-03-05 18:16 ` Jens Axboe
2007-03-01 23:36 ` Linus Torvalds
2007-03-02 10:07 ` Pavel Machek
2007-03-05 8:42 ` Michael S. Tsirkin
2007-03-05 10:11 ` SATA resume slowness, e1000 MSI warning Ingo Molnar
2007-03-06 5:30 ` Jeff Garzik
2007-03-06 6:35 ` Kok, Auke
2007-03-06 9:04 ` Ingo Molnar
2007-03-06 15:34 ` Kok, Auke
2007-03-07 4:15 ` Eric W. Biederman
2007-03-07 16:31 ` Kok, Auke
2007-03-07 16:45 ` Kok, Auke
2007-03-07 19:28 ` Eric W. Biederman
2007-03-08 2:53 ` Andrew Morton
2007-03-08 6:58 ` Eric W. Biederman
2007-03-08 9:55 ` Jeff Garzik
2007-03-08 17:27 ` Eric W. Biederman
2007-03-08 19:58 ` [PATCH 0/2] Repair pci_restore_state when used with device resets Eric W. Biederman
2007-03-08 20:04 ` [PATCH 1/2] msi: Safer state caching Eric W. Biederman
2007-03-08 20:06 ` Eric W. Biederman [this message]
2007-03-12 22:46 ` [PATCH 2/2] pci: Repair pci_save/restore_state so we can restore one save many times Kok, Auke
2007-03-08 20:08 ` [PATCH 0/2] Repair pci_restore_state when used with device resets Ingo Molnar
2007-03-08 20:26 ` Eric W. Biederman
2007-03-08 10:23 ` SATA resume slowness, e1000 MSI warning Michael S. Tsirkin
2007-03-11 11:11 ` Eric W. Biederman
2007-03-11 11:24 ` Michael S. Tsirkin
2007-03-11 17:37 ` Eric W. Biederman
2007-03-11 18:03 ` Michael S. Tsirkin
2007-03-11 18:27 ` Eric W. Biederman
2007-03-11 18:37 ` Michael S. Tsirkin
2007-03-11 19:50 ` Eric W. Biederman
2007-03-12 4:35 ` Michael S. Tsirkin
2007-04-16 19:56 ` Michael S. Tsirkin
2007-03-09 23:06 ` Kok, Auke
2007-03-10 3:41 ` Eric W. Biederman
2007-03-06 9:06 ` Ingo Molnar
2007-03-06 16:26 ` Thomas Gleixner
2007-03-06 16:52 ` Linus Torvalds
2007-03-06 17:09 ` Kok, Auke
2007-03-09 6:44 ` 2.6.21-rc1: known regressions (part 2) Pavel Machek
2007-03-05 15:34 ` Michael S. Tsirkin
2007-02-27 22:09 ` Adrian Bunk
2007-02-28 7:41 ` Jens Axboe
2007-02-26 22:01 ` 2.6.21-rc1: known regressions (v2) (part 1) Adrian Bunk
2007-02-27 4:09 ` Sergio Monteiro Basto
2007-02-27 12:50 ` S.Çağlar Onur
2007-02-27 13:25 ` Ismail Dönmez
2007-02-28 21:13 ` Michael S. Tsirkin
2007-02-28 21:27 ` Thomas Gleixner
2007-02-28 21:40 ` Michael S. Tsirkin
2007-03-01 3:45 ` Jeff Chua
2007-03-02 12:26 ` [linux-pm] " Pavel Machek
2007-03-03 11:17 ` Jens Axboe
2007-03-05 0:04 ` Adrian Bunk
2007-03-06 1:32 ` Jeff Chua
2007-03-06 12:03 ` Jeff Chua
2007-03-06 12:08 ` Michael S. Tsirkin
2007-03-06 12:12 ` Jeff Chua
2007-03-19 15:32 ` Pavel Machek
2007-03-19 21:23 ` Rafael J. Wysocki
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=m1lki75wlm.fsf_-_@ebiederm.dsl.xmission.com \
--to=ebiederm@xmission.com \
--cc=akpm@linux-foundation.org \
--cc=auke-jan.h.kok@intel.com \
--cc=bunk@stusta.de \
--cc=gregkh@suse.de \
--cc=jeff@garzik.org \
--cc=jens.axboe@oracle.com \
--cc=linux-kernel@vger.kernel.org \
--cc=linux-pci@atrey.karlin.mff.cuni.cz \
--cc=linux-pm@lists.osdl.org \
--cc=michael@ellerman.id.au \
--cc=michal.k.k.piotrowski@gmail.com \
--cc=mingo@elte.hu \
--cc=mst@mellanox.co.il \
--cc=pavel@ucw.cz \
--cc=tglx@linutronix.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