| lists.openwall.net | lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC | |
|
Open Source and information security mailing list archives
| ||
|
Message-ID: <001e01d859dc$3d4d6870$b7e83950$@samsung.com>
Date: Wed, 27 Apr 2022 11:12:26 +0900
From: 이범우 <bw365.lee@...sung.com>
To: "'Chanwoo Choi'" <cwchoi00@...il.com>,
"'MyungJoo Ham'" <myungjoo.ham@...sung.com>,
"'Chanwoo Choi'" <cw00.choi@...sung.com>,
<linux-kernel@...r.kernel.org>
Subject: RE: [PATCH] extcon: modify extcon device to be created after driver
data is set
Hi chanwoo.
Thank you for your review.
Actually, I thought the same thing as you first.
However, it was a problem because dev->devres_lock was not initialized and devm_kcalloc was used, and this patch was created instead of using kcalloc instead of devm_kcalloc.
In your opinion, if I use kcalloc instead of devm_kcalloc, there will be no problem.
But, I think the code of the following part should be supplemented.
1. alloc edev->nh when edev->max_supported is none zero.
@@ -1224,19 +1224,14 @@ int extcon_dev_register(struct extcon_dev *edev)
+ if (edev->max_supported) {
+ edev->nh = kcalloc(edev->max_supported, sizeof(*edev->nh),
+ GFP_KERNEL);
+ if (!edev->nh) {
+ ret = -ENOMEM;
+ goto err_alloc_nh;
+ }
}
2. free evdev->nh when device unregister.
@@ -1314,6 +1318,7 @@ void extcon_dev_unregister(struct extcon_dev *edev)
if (edev->max_supported) {
kfree(edev->extcon_dev_type.groups);
kfree(edev->cables);
+ kfree(edev->nh);
}
I will re-raise commit after testing.
Please let me know if you have additional concern.
Best Regards,
Bumwoo Lee.
-----Original Message-----
From: Chanwoo Choi <cwchoi00@...il.com>
Sent: Wednesday, April 27, 2022 2:09 AM
To: bumwoo lee <bw365.lee@...sung.com>; MyungJoo Ham <myungjoo.ham@...sung.com>; Chanwoo Choi <cw00.choi@...sung.com>; linux-kernel@...r.kernel.org
Subject: Re: [PATCH] extcon: modify extcon device to be created after driver data is set
Hi Bumwoo,
Firstly, I'm sorry for late replay.
On 22. 3. 31. 12:03, bumwoo lee wrote:
> Currently, someone can invoke the sysfs such as state_show()
> intermittently before set_drvdata() is done.
> And it can be a cause of kernel Oops because of edev is Null at that time.
> So modified the driver registration to after setting drviver data.
>
> - Oops's backtrace.
>
> Backtrace:
> [<c067865c>] (state_show) from [<c05222e8>] (dev_attr_show)
> [<c05222c0>] (dev_attr_show) from [<c02c66e0>] (sysfs_kf_seq_show)
> [<c02c6648>] (sysfs_kf_seq_show) from [<c02c496c>] (kernfs_seq_show)
> [<c02c4938>] (kernfs_seq_show) from [<c025e2a0>] (seq_read)
> [<c025e11c>] (seq_read) from [<c02c50a0>] (kernfs_fop_read)
> [<c02c5064>] (kernfs_fop_read) from [<c0231cac>] (__vfs_read)
> [<c0231c5c>] (__vfs_read) from [<c0231ee0>] (vfs_read) [<c0231e34>]
> (vfs_read) from [<c0232464>] (ksys_read) [<c02323f0>] (ksys_read) from
> [<c02324fc>] (sys_read) [<c02324e4>] (sys_read) from [<c00091d0>]
> (__sys_trace_return)
>
> Signed-off-by: bumwoo lee <bw365.lee@...sung.com>
> ---
> drivers/extcon/extcon.c | 12 ++++++------
> 1 file changed, 6 insertions(+), 6 deletions(-)
>
> diff --git a/drivers/extcon/extcon.c b/drivers/extcon/extcon.c index
> d27cd9f88f86..74fee04bd764 100644
> --- a/drivers/extcon/extcon.c
> +++ b/drivers/extcon/extcon.c
> @@ -1224,18 +1224,13 @@ int extcon_dev_register(struct extcon_dev *edev)
> edev->dev.type = &edev->extcon_dev_type;
> }
>
> - ret = device_register(&edev->dev);
> - if (ret) {
> - put_device(&edev->dev);
> - goto err_dev;
> - }
> + device_initialize(&edev->dev);
>
> spin_lock_init(&edev->lock);
> edev->nh = devm_kcalloc(&edev->dev, edev->max_supported,
> sizeof(*edev->nh), GFP_KERNEL);
Actually, I'm not sure that it is proper to use 'edev->dev'
by devm_kcalloc before device_add. When error happen between device_initialize and device_add, I think that it is not properly handled such as free the allocated memory automatically.
> if (!edev->nh) {
> ret = -ENOMEM;
> - device_unregister(&edev->dev);
> goto err_dev;
> }
>
> @@ -1251,9 +1246,14 @@ int extcon_dev_register(struct extcon_dev *edev)
> list_add(&edev->entry, &extcon_dev_list);
> mutex_unlock(&extcon_dev_list_lock);
>
> + ret = device_add(&edev->dev);
> + if (ret)
> + goto err_dev;
On this case, need to delete entry from extcon_dev_list by using list_del.
> +
> return 0;
>
> err_dev:
> + put_device(&edev->dev);
> if (edev->max_supported)
> kfree(edev->extcon_dev_type.groups);
> err_alloc_groups:
I recommend that move device_register() position after dev_set_drvdata.
And then use kcalloc instead of devm_kcalloc as following:
diff --git a/drivers/extcon/extcon.c b/drivers/extcon/extcon.c index d27cd9f88f86..80e5bfec1131 100644
--- a/drivers/extcon/extcon.c
+++ b/drivers/extcon/extcon.c
@@ -1224,19 +1224,11 @@ int extcon_dev_register(struct extcon_dev *edev)
edev->dev.type = &edev->extcon_dev_type;
}
- ret = device_register(&edev->dev);
- if (ret) {
- put_device(&edev->dev);
- goto err_dev;
- }
-
spin_lock_init(&edev->lock);
- edev->nh = devm_kcalloc(&edev->dev, edev->max_supported,
- sizeof(*edev->nh), GFP_KERNEL);
+ edev->nh = kcalloc(edev->max_supported, sizeof(*edev->nh),
GFP_KERNEL);
if (!edev->nh) {
ret = -ENOMEM;
- device_unregister(&edev->dev);
- goto err_dev;
+ goto err_alloc_nh;
}
for (index = 0; index < edev->max_supported; index++) @@ -1247,6 +1239,12 @@ int extcon_dev_register(struct extcon_dev *edev)
dev_set_drvdata(&edev->dev, edev);
edev->state = 0;
+ ret = device_register(&edev->dev);
+ if (ret) {
+ put_device(&edev->dev);
+ goto err_dev;
+ }
+
mutex_lock(&extcon_dev_list_lock);
list_add(&edev->entry, &extcon_dev_list);
mutex_unlock(&extcon_dev_list_lock);
@@ -1254,6 +1252,9 @@ int extcon_dev_register(struct extcon_dev *edev)
return 0;
err_dev:
+ if (edev->max_supported)
+ kfree(edev->nh);
+err_alloc_nh:
if (edev->max_supported)
kfree(edev->extcon_dev_type.groups);
err_alloc_groups:
@@ -1314,6 +1315,7 @@ void extcon_dev_unregister(struct extcon_dev *edev)
if (edev->max_supported) {
kfree(edev->extcon_dev_type.groups);
kfree(edev->cables);
+ kfree(edev->nh);
}
put_device(&edev->dev);
--
Best Regards,
Samsung Electronics
Chanwoo Choi
Powered by blists - more mailing lists