netlink: protection and workaround for too long mcast group name
Subsystems (4)
cover netlink: protection and workaround for too long mcast group name
You will see garbage at the end of line in the output of following command line:
$ genl ctrl show | grep thermal_mc_group
#1: ID-0x2 name: thermal_mc_group^BThe type of structure field for "name" is char[16]:
#define GENL_NAMSIZ 16 /* length of family name */
...
struct genl_multicast_group {
...
char name[GENL_NAMSIZ];
...
};strlen("thermal_mc_group") == 16 is too long for the array size.
This patch series provid a protection(patch for genetlink) for this kind of bug and workaround(patch for thermal).
Masatake YAMATO (2):
genetlink: trigger BUG_ON if a group name is too long
thermal: shorten too long mcast group name
include/linux/thermal.h | 2 +- net/netlink/genetlink.c | 1 + 2 files changed, 2 insertions(+), 1 deletion(-)
Signature · 1 line
1.7.11.7
PATCH 1/2 genetlink: trigger BUG_ON if a group name is too long
Trigger BUG_ON if a group name is longer than GENL_NAMSIZ.
--- net/netlink/genetlink.c | 1 + 1 file changed, 1 insertion(+)
net/netlink/genetlink.c +1 -0
@@ -142,6 +142,7 @@ int genl_register_mc_group(struct genl_family *family,int err =0 ;BUG_ON (grp->name[0 ] =='\0' ); +BUG_ON (memchr (grp->name,'\0' ,GENL_NAMSIZ ) ==NULL );genl_lock ();
Signature · 1 line
1.7.11.7
Re: [PATCH 1/2] genetlink: trigger BUG_ON if a group name is too long
From: Masatake YAMATO <[email protected]>
Date: Tue, 19 Mar 2013 20:47:27 +0900
Quoted 3 lines
Trigger
BUG_ONif a group name is longer thanGENL_NAMSIZ.Signed-off-by Masatake YAMATO <[email protected]>
Applied, thanks.
Although I'm disappointed that the compiler doesn't say anything about this in the assignment.
We're assigning "char[16] = STRING" and it doesn't say anything if the final NULL char doesn't fit into the array.
Re: [PATCH 1/2] genetlink: trigger BUG_ON if a group name is too long
Quoted 7 lines · David Miller
From: Masatake YAMATO <[email protected]>
Date: Tue, 19 Mar 2013 20:47:27 +0900Quoted 3 lines
Trigger
BUG_ONif a group name is longer thanGENL_NAMSIZ.Signed-off-by Masatake YAMATO <[email protected]>
Quoted 7 lines
Applied, thanks.
Although I'm disappointed that the compiler doesn't say anything about this in the assignment.
We're assigning "char[16] = STRING" and it doesn't say anything if the final NULL char doesn't fit into the array.
Unfortunately the C standard says this is allowed. (Though the C++ standard says it's not!) That doesn't mean the compiler can't warn about it, of course.
Ben.
Signature · 3 lines
Ben Hutchings, Staff Engineer, Solarflare Not speaking for my employer; that's the marketing department's job. They asked us to note that Solarflare product names are trademarked.
Re: [PATCH 1/2] genetlink: trigger BUG_ON if a group name is too long
I will look at sparse whether it can capture this kind of mistake.
Masatake YAMATO
From: Ben Hutchings <[email protected]>
Date: Wed, 20 Mar 2013 17:32:59 +0000
Quoted 27 lines · David Miller
Quoted 7 lines
From: Masatake YAMATO <[email protected]>
Date: Tue, 19 Mar 2013 20:47:27 +0900Quoted 3 lines
Trigger
BUG_ONif a group name is longer thanGENL_NAMSIZ.Signed-off-by Masatake YAMATO <[email protected]>Quoted 7 lines
Applied, thanks.
Although I'm disappointed that the compiler doesn't say anything about this in the assignment.
We're assigning "char[16] = STRING" and it doesn't say anything if the final NULL char doesn't fit into the array.
Unfortunately the C standard says this is allowed. (Though the C++ standard says it's not!) That doesn't mean the compiler can't warn about it, of course.
Ben.
Signature · 3 lines
Ben Hutchings, Staff Engineer, Solarflare Not speaking for my employer; that's the marketing department's job. They asked us to note that Solarflare product names are trademarked.
Re: [PATCH 1/2] genetlink: trigger BUG_ON if a group name is too long
In reply to David MillerQuoted 8 lines · David Miller
From: Masatake YAMATO <[email protected]>
Date: Tue, 19 Mar 2013 20:47:27 +0900Quoted 3 lines
Trigger
BUG_ONif a group name is longer thanGENL_NAMSIZ.Signed-off-by Masatake YAMATO <[email protected]>Applied, thanks.
catch one in
drivers/thermal/thermal_sys.c::genetlink_init
result = genl_register_mc_group(&thermal_event_genl_family,
&thermal_event_mcgrp);
and
static struct genl_multicast_group thermal_event_mcgrp = {
.name = THERMAL_GENL_MCAST_GROUP_NAME,
};
#define THERMAL_GENL_MCAST_GROUP_NAME "thermal_mc_group"
that string len 16.
David Miller
maintainerRe: [PATCH 1/2] genetlink: trigger BUG_ON if a group name is too long
From: Yinghai Lu <[email protected]>
Date: Wed, 20 Mar 2013 14:01:09 -0700
Quoted 11 lines · David Miller
Quoted 8 lines
From: Masatake YAMATO <[email protected]>
Date: Tue, 19 Mar 2013 20:47:27 +0900Quoted 3 lines
Trigger
BUG_ONif a group name is longer thanGENL_NAMSIZ.Signed-off-by Masatake YAMATO <[email protected]>Applied, thanks.
catch one in
Seriously?
That's what his second patch in this series fixes, we know about it already.
Re: [PATCH 1/2] genetlink: trigger BUG_ON if a group name is too long
Quoted 11 lines · David Miller
From: Yinghai Lu <[email protected]>
Date: Wed, 20 Mar 2013 14:01:09 -0700Quoted 8 lines · David Miller
Quoted 7 lines
From: Masatake YAMATO <[email protected]>
Date: Tue, 19 Mar 2013 20:47:27 +0900Quoted 3 lines
Trigger
BUG_ONif a group name is longer thanGENL_NAMSIZ.Signed-off-by Masatake YAMATO <[email protected]>
Quoted 8 lines
Quoted 3 lines
Quoted 1 line
Applied, thanks.
catch one in
Seriously?
That's what his second patch in this series fixes, we know about it already.
[ 6.047966] ------------[ cut here ]------------ [ 6.048944] kernel BUG at net/netlink/genetlink.c:145! [ 6.049204] invalid opcode: 0000 [#1] SMP [ 6.049501] Modules linked in: [ 6.049728] CPU 2 [ 6.050003] Pid: 1, comm: swapper/0 Not tainted 3.9.0-rc3-yh-00486-gb6d1e23-dirty #1371 Bochs Bochs [ 6.050299] RIP: 0010:[<ffffffff81f0f968>] [<ffffffff81f0f968>] genl_register_mc_group+0x38/0x270 [ 6.050593] RSP: 0000:ffff880119e63dd8 EFLAGS: 00000246 [ 6.050804] RAX: 0000000000000000 RBX: 0000000000000000 RCX: ffffffff82b667c8 [ 6.050931] RDX: ffffffff82b667c8 RSI: 0000000000000000 RDI: ffffffff82b667b8 [ 6.050931] RBP: ffff880119e63e28 R08: 0000000000000000 R09: 0000000000000000 [ 6.050931] R10: 0000000000000000 R11: 0000000000000001 R12: ffffffff82b667a0 [ 6.050931] R13: ffffffff82b66720 R14: 00000001683fe47c R15: 0000000000000000 [ 6.050931] FS: 0000000000000000(0000) GS:ffff88019f600000(0000) knlGS:0000000000000000 [ 6.050931] CS: 0010 DS: 0000 ES: 0000 CR0: 000000008005003b [ 6.050931] CR2: 0000000000000000 CR3: 0000000002a14000 CR4: 00000000000006e0 [ 6.050931] DR0: 0000000000000000 DR1: 0000000000000000 DR2: 0000000000000000 [ 6.050931] DR3: 0000000000000000 DR6: 0000000000000000 DR7: 0000000000000000 [ 6.050931] Process swapper/0 (pid: 1, threadinfo ffff880119e62000, task ffff880119e68000) [ 6.050931] Stack: [ 6.050931] ffff880119e63de8 ffffffff82b66720 ffffffff82dcb082 0000000000000000 [ 6.050931] ffff880119e63e28 ffffffff81f0f901 0000000000000000 ffffffff82dcb082 [ 6.050931] 0000000000000000 00000001683fe47c ffff880119e63e48 ffffffff82dcb0f8 [ 6.050931] Call Trace: [ 6.050931] [<ffffffff82dcb082>] ? smsc47b397_init+0x19d/0x19d [ 6.050931] [<ffffffff81f0f901>] ? genl_register_family+0x1b1/0x1e0 [ 6.050931] [<ffffffff82dcb082>] ? smsc47b397_init+0x19d/0x19d [ 6.050931] [<ffffffff82dcb0f8>] thermal_init+0x76/0x8e [ 6.050931] [<ffffffff81000254>] do_one_initcall+0x64/0x160 [ 6.050931] [<ffffffff82d8a321>] kernel_init_freeable+0x1bf/0x251 [ 6.050931] [<ffffffff82d89a71>] ? loglevel+0x31/0x31 [ 6.050931] [<ffffffff81ff0740>] ? rest_init+0xc0/0xc0 [ 6.050931] [<ffffffff81ff074e>] kernel_init+0xe/0xf0 [ 6.050931] [<ffffffff8202999c>] ret_from_fork+0x7c/0xb0 [ 6.050931] [<ffffffff81ff0740>] ? rest_init+0xc0/0xc0 [ 6.050931] Code: 41 54 49 89 f4 53 48 83 ec 30 80 7e 18 00 75 03 0f 0b 90 49 89 fd 48 8d 7e 18 ba 10 00 00 00 31 f6 e8 3d 01 5e ff 48 85 c0 75 08 <0f> 0b 66 0f 1f 44 00 00 e8 7b f3 ff ff 49 81 fc a0 20 b8 82 74 [ 6.050931] RIP [<ffffffff81f0f968>] genl_register_mc_group+0x38/0x270 [ 6.050931] RSP <ffff880119e63dd8> [ 6.063404] ---[ end trace 60aaac53e4d8e418 ]---
David Miller
maintainerRe: [PATCH 1/2] genetlink: trigger BUG_ON if a group name is too long
From: Yinghai Lu <[email protected]>
Date: Wed, 20 Mar 2013 14:06:06 -0700
Quoted 16 lines · David Miller
Quoted 15 lines
From: Yinghai Lu <[email protected]>
Date: Wed, 20 Mar 2013 14:01:09 -0700Quoted 11 lines · David Miller
Quoted 8 lines
From: Masatake YAMATO <[email protected]>
Date: Tue, 19 Mar 2013 20:47:27 +0900Quoted 3 lines
Trigger
BUG_ONif a group name is longer thanGENL_NAMSIZ.Signed-off-by Masatake YAMATO <[email protected]>Applied, thanks.
catch one in
Quoted 7 lines
Quoted 4 lines
Seriously?
That's what his second patch in this series fixes, we know about it already.
[ 6.047966] ------------[ cut here ]------------ [ 6.048944] kernel BUG at net/netlink/genetlink.c:145!
For the second time, we know about this bug.
His second patch fixes the themal layer bug.
But I told him to send it to the thermal maintainers so that they can integrate it.
Why are you posting what we know already over and over?
Re: [PATCH 1/2] genetlink: trigger BUG_ON if a group name is too long
Quoted 4 lines · David Miller
Quoted 3 lines
[ 6.047966] ------------[ cut here ]------------ [ 6.048944] kernel BUG at net/netlink/genetlink.c:145!
Quoted 8 lines
For the second time, we know about this bug.
His second patch fixes the themal layer bug.
But I told him to send it to the thermal maintainers so that they can integrate it.
Why are you posting what we know already over and over?
so bisecting will be broken?
David Miller
maintainerRe: [PATCH 1/2] genetlink: trigger BUG_ON if a group name is too long
From: Yinghai Lu <[email protected]>
Date: Wed, 20 Mar 2013 14:16:22 -0700
Quoted 7 lines · David Miller
Quoted 6 lines
Quoted 3 lines
[ 6.047966] ------------[ cut here ]------------ [ 6.048944] kernel BUG at net/netlink/genetlink.c:145!For the second time, we know about this bug.
Quoted 8 lines
Quoted 6 lines
His second patch fixes the themal layer bug.
But I told him to send it to the thermal maintainers so that they can integrate it.
Why are you posting what we know already over and over?
so bisecting will be broken?
Good point, so I'll apply the thermal patch to my tree to minimize the failure range.
Thanks.
Re: [PATCH 1/2] genetlink: trigger BUG_ON if a group name is too long
Quoted 4 lines · David Miller
Quoted 1 line
so bisecting will be broken?
Good point, so I'll apply the thermal patch to my tree to minimize the failure range.
yes, that "second" patch should be applied before this one.
David Miller
maintainerRe: [PATCH 1/2] genetlink: trigger BUG_ON if a group name is too long
From: Yinghai Lu <[email protected]>
Date: Wed, 20 Mar 2013 15:17:13 -0700
Quoted 8 lines · David Miller
Quoted 5 lines
Quoted 2 lines
so bisecting will be broken?
Good point, so I'll apply the thermal patch to my tree to minimize the failure range.
yes, that "second" patch should be applied before this one.
But it's too late as the first one is already commited to my public GIT tree on kernel.org and I never rebase my tree, ever.
So this is the only option to minimize the damage at this point.
Re: [PATCH 1/2] genetlink: trigger BUG_ON if a group name is too long
In reply to David MillerFrom: David Miller <[email protected]>
Subject: Re: [PATCH 1/2] genetlink: trigger BUG_ON if a group name is too long
Date: Wed, 20 Mar 2013 12:07:05 -0400 (EDT)
Quoted 7 lines
From: Masatake YAMATO <[email protected]>
Date: Tue, 19 Mar 2013 20:47:27 +0900Quoted 3 lines
Trigger
BUG_ONif a group name is longer thanGENL_NAMSIZ.Signed-off-by Masatake YAMATO <[email protected]>
Quoted 7 lines
Applied, thanks.
Although I'm disappointed that the compiler doesn't say anything about this in the assignment.
We're assigning "char[16] = STRING" and it doesn't say anything if the final NULL char doesn't fit into the array.
Now sparse with -Winit-cstring option can detect this kind assignment.
Masatake YAMATO
PATCH 2/2 thermal: shorten too long mcast group name
In reply to Masatake YAMATOThe original name is too long.
--- include/linux/thermal.h | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-)
include/linux/thermal.h +1 -1
@@ -44,7 +44,7 @@/* Adding event notification support elements */ #define THERMAL_GENL_FAMILY_NAME "thermal_event"#define THERMAL_GENL_VERSION 0x01 -#define THERMAL_GENL_MCAST_GROUP_NAME "thermal_mc_group" +#define THERMAL_GENL_MCAST_GROUP_NAME "thermal_mc_grp"/* Default Thermal Governor */ #if defined(CONFIG_THERMAL_DEFAULT_GOV_STEP_WISE )
Signature · 1 line
1.7.11.7
David Miller
maintainerRe: [PATCH 2/2] thermal: shorten too long mcast group name
From: Masatake YAMATO <[email protected]>
Date: Tue, 19 Mar 2013 20:47:28 +0900
Quoted 3 lines
The original name is too long.
Signed-off-by Masatake YAMATO <[email protected]>
This change needs to go in via the Thermal driver maintainers.
Re: [PATCH 2/2] thermal: shorten too long mcast group name
Quoted 8 lines
From: Masatake YAMATO <[email protected]>
Date: Tue, 19 Mar 2013 20:47:28 +0900Quoted 3 lines
The original name is too long.
Signed-off-by Masatake YAMATO <[email protected]>This change needs to go in via the Thermal driver maintainers.
Thanks. I've submitted the patch to linux-acpi ML.
Masatake YAMATO
David Miller
maintainerRe: [PATCH 2/2] thermal: shorten too long mcast group name
From: Masatake YAMATO <[email protected]>
Date: Thu, 21 Mar 2013 16:46:23 +0900 (JST)
Quoted 10 lines
Quoted 8 lines
From: Masatake YAMATO <[email protected]>
Date: Tue, 19 Mar 2013 20:47:28 +0900Quoted 3 lines
The original name is too long.
Signed-off-by Masatake YAMATO <[email protected]>This change needs to go in via the Thermal driver maintainers.
Thanks. I've submitted the patch to linux-acpi ML.
I already applied the patch to my tree, didn't you read the rest of this thread?
And in the future, order your patches such that you fix the bug before adding the assertion.
Re: [PATCH 2/2] thermal: shorten too long mcast group name
Sorry for cross posting to netdev and linux-acpi.
I submitted almost same patch about thermal subsystem both netdev list and then linux-acpi list.
http://marc.info/?l=linux-netdev&m=136371412725264&w=2
http://marc.info/?l=linux-acpi&m=136386050309109&w=2
After submitting to linux-acpi list, I noticed the patch is merged to a tree davem maintained.
What I should do?
The submitted pathes are bit different.
In netdev, I proposed changing "thermal_mc_group" to "thermal_mc_grp". In linux-acpi I proposed changing "thermal_mc_group" to "thermal_mc_grou".
"thermal_mc_grp" is easier to read. However, when thinking about compatibility, "thermal_mc_grou" is better.
I've already submitted another patch for genl command which truncates the C string returned from kernel.
http://marc.info/?l=linux-netdev&m=136371412625262&w=2
Consider a situation(A); using old kernel(without my patch submitted for linux-acpi) and new genl command (with my patch for genl). You will get
thermal_mc_grou
as output.
Consider anther situation(B); using new kernel(with my patch submitted for linux-acpi) and old genl command (without my patch for genl). You will get
thermal_mc_grou
as output. In both situations(A and B) users can get the same output. If you choose "thermal_mc_grp", users may see "thermal_mc_grp" or "thermal_mc_grou".
Quoted 5 lines
Quoted 5 lines
Quoted 5 lines
From: Masatake YAMATO <[email protected]>
Date: Tue, 19 Mar 2013 20:47:28 +0900Quoted 2 lines
The original name is too long.
Quoted 8 lines
Quoted 5 lines
Quoted 3 lines
Quoted 1 line
Signed-off-by Masatake YAMATO <[email protected]>This change needs to go in via the Thermal driver maintainers.
Thanks. I've submitted the patch to linux-acpi ML.
I already applied the patch to my tree, didn't you read the rest of this thread?
I read but I didn't understand the meaning of applying because I cannot find my 2nd patch for thermal in net-next tree when I pulled.
What I should do? Should I withdraw the patch submitted to linux-acpi?
However, I slight modified the patch when I submitted it to linux-acpi.
Quoted 2 lines
And in the future, order your patches such that you fix the bug before adding the assertion.
Sorry. I'll do so next time.
Masatake YAMATO
David Miller
maintainerRe: [PATCH 2/2] thermal: shorten too long mcast group name
From: Masatake YAMATO <[email protected]>
Date: Fri, 22 Mar 2013 12:43:06 +0900 (JST)
Quoted 3 lines
I read but I didn't understand the meaning of applying because I cannot find my 2nd patch for thermal in net-next tree when I pulled.
These are bug fixes, so they appear in the 'net' tree not 'net-next'.
Re: [PATCH 2/2] thermal: shorten too long mcast group name
In reply to Masatake YAMATOQuoted 12 lines · Masatake YAMATO
The original name is too long.
Signed-off-by Masatake YAMATO <[email protected]>--- include/linux/thermal.h | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-)include/linux/thermal.h +0
-0
include/linux/thermal.h line 47
Quoted 5 lines
include/linux/thermal.h +1
-1@@ -44,7 +44,7 @@/* Adding event notification support elements */ #define THERMAL_GENL_FAMILY_NAME "thermal_event"#define THERMAL_GENL_VERSION 0x01 -#define THERMAL_GENL_MCAST_GROUP_NAME "thermal_mc_group" +#define THERMAL_GENL_MCAST_GROUP_NAME "thermal_mc_grp"
Just saying that this would not be OK in general as this would break any application looking up the multicast group id by name via genetlink but since it's broken anyway right now it's OK I guess.
PATCH 1/1 thermal: shorten too long mcast group name
You will see garbage at the end of line in the output of following command line:
$ genl ctrl show | grep thermal_mc_group
#1: ID-0x2 name: thermal_mc_group^BThe type of structure field for "name" is char[16] in the kernel:
#define GENL_NAMSIZ 16 /* length of family name */
...
struct genl_multicast_group {
...
char name[GENL_NAMSIZ];
...
};strlen("thermal_mc_group") == 16 is too long for the array size.
"thermal_mc_grp" is more readable than "thermal_mc_grou". However, I choose latter; the patch for genl command side may truncate the last char in the name:
http://marc.info/?l=linux-netdev&m=136371412625262&w=2
--- include/linux/thermal.h | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-)
include/linux/thermal.h +1 -1
@@ -44,7 +44,7 @@/* Adding event notification support elements */ #define THERMAL_GENL_FAMILY_NAME "thermal_event"#define THERMAL_GENL_VERSION 0x01 -#define THERMAL_GENL_MCAST_GROUP_NAME "thermal_mc_group" +#define THERMAL_GENL_MCAST_GROUP_NAME "thermal_mc_grou"/* Default Thermal Governor */ #if defined(CONFIG_THERMAL_DEFAULT_GOV_STEP_WISE )
Signature · 1 line
1.7.11.7
RE: [PATCH] thermal: shorten too long mcast group name
Quoted 8 lines
-----Original Message-----
From: [email protected] [mailto:linux-acpi- [email protected]] On Behalf Of Masatake YAMATO
Sent: Thursday, March 21, 2013 3:38 PM
To: [email protected]
Cc: [email protected]
Subject: [PATCH] thermal: shorten too long mcast group name
Quoted 5 lines
You will see garbage at the end of line in the output of following command line:
$ genl ctrl show | grep thermal_mc_group #1: ID-0x2 name: thermal_mc_group^B
Agreed.
Quoted 11 lines
The type of structure field for "name" is char[16] in the kernel:
#define GENL_NAMSIZ 16 /* length of family name */ ... struct genl_multicast_group { ... char name[GENL_NAMSIZ]; ... };
Quoted 5 lines
strlen("thermal_mc_group")== 16 is too long for the array size."
thermal_mc_grp" is more readable than "thermal_mc_grou". However, I choose latter; the patch for genl command side may truncate the last char in the name:
I prefer we use thermal_mc_grp.
+ Adding Rui & the Linux-pm list.
Thanks,
Durga
Quoted 29 lines
http://marc.info/?l=linux-netdev&m=136371412625262&w=2Signed-off-by Masatake YAMATO <[email protected]>--- include/linux/thermal.h | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-)include/linux/thermal.h +0
-1@@ -44,7 +44,7 @@/* Adding event notification support elements */ #define THERMAL_GENL_FAMILY_NAME "thermal_event"#define THERMAL_GENL_VERSION 0x01 -#define THERMAL_GENL_MCAST_GROUP_NAME "
thermal_mc_group"
+#defineTHERMAL_GENL_MCAST_GROUP_NAME"thermal_mc_grou"/* Default Thermal Governor */
#ifdefined(CONFIG_THERMAL_DEFAULT_GOV_STEP_WISE)
--
1.7.11.7--
To unsubscribe from this list: send the line "unsubscribe linux-acpi" in the body of a message to [email protected]
More majordomo info at http://vger.kernel.org/majordomo-info.html