Skip to contentSkip to conversation

+ fix-compat-space-msg-size-limit-for-msgsnd-msgrcv.patch added to -mm tree

DISCUSSION Arnd Bergmann 2 participants 6 messages Nov 14, 2006 1 message not in the archive

lore
Share & tools
lore

A message that is not in the archive (sent off-list or not yet received): [email protected]

#1 lore raw
Quoted 12 lines · On Tuesday 14 November 2006 00:58, [email protected] wrote:

Subject: Fix compat space msg size limit for msgsnd/msgrcv
From: suzuki <[email protected]>

Currently we allocate 64k space on the user stack and use it the msgbuf for sys_{msgrcv,msgsnd} for compat and the results are later copied in user [by copy_in_user].

This patch introduces helper routines for sys_{msgrcv,msgsnd} which would accept the pointer to msgbuf along with the msgp->mtext. This avoids the need to allocate the msgsize on the userspace (thus removing the size limit) and the overhead of an extra copy_in_user().

Quoted 4 lines
Cc Arnd Bergmann <[email protected]>
Cc "David S. Miller" <[email protected]>

This patch is definitely a big step in the right direction here, but why not go all the way and pass msgp->mtype to do_msgsnd/do_msgrcv as kernel data instead of a user space pointer? This way you can get rid of the compat_alloc_userspace entirely and save avoid doing an extra put_user/get_user pair in the compat_ function.

	Arnd <><
#2 lore raw
Quoted 22 lines · Arnd Bergmann
Quoted 12 lines · On Tuesday 14 November 2006 00:58, [email protected] wrote:

Subject: Fix compat space msg size limit for msgsnd/msgrcv
From: suzuki <[email protected]>

Currently we allocate 64k space on the user stack and use it the msgbuf for sys_{msgrcv,msgsnd} for compat and the results are later copied in user [by copy_in_user].

This patch introduces helper routines for sys_{msgrcv,msgsnd} which would accept the pointer to msgbuf along with the msgp->mtext. This avoids the need to allocate the msgsize on the userspace (thus removing the size limit) and the overhead of an extra copy_in_user().

Quoted 4 lines
Cc Arnd Bergmann <[email protected]>
Cc "David S. Miller" <[email protected]>
Quoted 4 lines

This patch is definitely a big step in the right direction here, but why not go all the way and pass msgp->mtype to do_msgsnd/do_msgrcv as kernel data instead of a user space pointer? This way you can get rid of the
compat_alloc_userspace entirely and save avoid doing an extra put_user/get_user pair in the compat_ function.

I left it as such, inorder to avoid the future changes that may come in the struct msgbuf -if at all-, which would make us to pass every single field as a parameter to do_msgrcv/do_msgsnd.

thanks,

Suzuki

Quoted 1 line
	Arnd <><
#3 lore raw
Quoted 4 lines · suzuki

I left it as such, inorder to avoid the future changes that may come in the struct msgbuf -if at all-, which would make us to pass every single field as a parameter to do_msgrcv/do_msgsnd.

struct msgbuf is part of the kernel ABI and will never change, so that's no problem at all.

	Arnd <><
#4 lore raw

Hello Arnd,

Quoted 7 lines · Arnd Bergmann
Quoted 1 line · suzuki

I left it as such, inorder to avoid the future changes that may come in the struct msgbuf -if at all-, which would make us to pass every single field as a parameter to do_msgrcv/do_msgsnd.

struct msgbuf is part of the kernel ABI and will never change, so that's no problem at all.

Ok.

Does the following change look fine ?

do_msgsnd() - Accepting the mtype and user space ptr to the mtext. i.e.,

long do_msgsnd(int msqid, long mtype, void __user *mtext,

		size_t msgsz, int msgflg);

and,

do_msgrcv() - accepting the kernel space data ptr to pmtype and user space ptr to mtext. The caller has to copy the *pmtype back to the user space.

i.e.,

long do_msgrcv(int msqid, long *pmtype, void __user *mtext,

                       size_t msgsz, long msgtyp, int msgflg);

or

Can we use the kernel space "struct msgbuf" instead of the mtype being passed explicitly.

Thanks,

Suzuki

Quoted 2 lines
	Arnd <><
#5 lore raw
Quoted 8 lines · suzuki

Does the following change look fine ?

do_msgsnd() - Accepting the mtype and user space ptr to the mtext. i.e.,

long do_msgsnd(int msqid, long mtype, void __user *mtext, size_t msgsz, int msgflg);
and,

Quoted 8 lines

do_msgrcv() - accepting the kernel space data ptr to pmtype and user space ptr to mtext. The caller has to copy the *pmtype back to the user space.

i.e.,

long do_msgrcv(int msqid, long *pmtype, void __user *mtext, size_t msgsz, long msgtyp, int msgflg);

Yes, that looks fine.

Quoted 2 lines

Can we use the kernel space "struct msgbuf" instead of the mtype being passed explicitly.

That works as well, although it may be a little confusing to have the extra mtext byte of that structure included there, so I'd prefer the first solution.

	Arnd <><
#6 lore raw
Quoted 21 lines · Arnd Bergmann
Quoted 6 lines · suzuki

Does the following change look fine ?

do_msgsnd() - Accepting the mtype and user space ptr to the mtext. i.e.,

long do_msgsnd(int msqid, long mtype, void __user *mtext,

               size_t msgsz, int msgflg);
Quoted 8 lines

and,

do_msgrcv() - accepting the kernel space data ptr to pmtype and user space ptr to mtext. The caller has to copy the *pmtype back to the user space.

i.e.,

long do_msgrcv(int msqid, long *pmtype, void __user *mtext,

                       size_t msgsz, long msgtyp, int msgflg);

Yes, that looks fine.

Quoted 6 lines
Quoted 1 line

Can we use the kernel space "struct msgbuf" instead of the mtype being passed explicitly.

That works as well, although it may be a little confusing to have the extra mtext byte of that structure included there, so I'd prefer the first solution.

Arnd,

Thanks for the input !

Here we go ! As per the suggestions, I have modified the patch.

Thanks,

Suzuki

Quoted 2 lines
	Arnd <><

* Fix the size limit of compat space msgsize.

Currently we allocate 64k space on the user stack and use it the msgbuf for sys_{msgrcv,msgsnd} for compat and the results are later copied in user [ by copy_in_user]. This patch introduces helper routines for sys_{msgrcv,msgsnd} as below:

do_msgsnd() : Accepts the mtype and user space ptr to the buffer along with the msqid and msgflg.
do_msgrcv() : Accepts a kernel space ptr to mtype and a userspace ptr to the buffer. The mtype has to be copied back the user space msgbuf by the caller.

These changes avoid the need to allocate the msgsize on the userspace ( thus removing the size limt ) and the overhead of an extra copy_in_user().

include/linux/msg.h +6 -0
Index: linux-2.6.19-rc1/include/linux/msg.h
===================================================================
@@ -92,6 +92,12 @@
 	struct list_head q_senders;
 };
 
+/* Helper routines for sys_msgsnd and sys_msgrcv */
+extern long do_msgsnd(int msqid, long mtype, void __user *mtext,
+			size_t msgsz, int msgflg);
+extern long do_msgrcv(int msqid, long *pmtype, void __user *mtext,
+			size_t msgsz, long msgtyp, int msgflg);
+
 #endif /* __KERNEL__ */
 
 #endif /* _LINUX_MSG_H */
ipc/compat.c +7 -16
Index: linux-2.6.19-rc1/ipc/compat.c
===================================================================
@@ -115,7 +115,6 @@
 
 extern int sem_ctls[];
 #define sc_semopm	(sem_ctls[2])
-#define MAXBUF (64*1024)
 
 static inline int compat_ipc_parse_version(int *cmd)
 {
@@ -307,35 +306,30 @@
 
 long compat_sys_msgsnd(int first, int second, int third, void __user *uptr)
 {
-	struct msgbuf __user *p;
 	struct compat_msgbuf __user *up = uptr;
 	long type;
 
 	if (first < 0)
 		return -EINVAL;
-	if (second < 0 || (second >= MAXBUF - sizeof(struct msgbuf)))
+	if (second < 0)
 		return -EINVAL;
 
-	p = compat_alloc_user_space(second + sizeof(struct msgbuf));
-	if (get_user(type, &up->mtype) ||
-	    put_user(type, &p->mtype) ||
-	    copy_in_user(p->mtext, up->mtext, second))
+	if (get_user(type, &up->mtype)) 
 		return -EFAULT;
 
-	return sys_msgsnd(first, p, second, third);
+	return do_msgsnd(first, type, up->mtext, second, third);
 }
 
 long compat_sys_msgrcv(int first, int second, int msgtyp, int third,
 			   int version, void __user *uptr)
 {
-	struct msgbuf __user *p;
 	struct compat_msgbuf __user *up;
 	long type;
 	int err;
 
 	if (first < 0)
 		return -EINVAL;
-	if (second < 0 || (second >= MAXBUF - sizeof(struct msgbuf)))
+	if (second < 0)
 		return -EINVAL;
 
 	if (!version) {
@@ -349,14 +343,11 @@
 		uptr = compat_ptr(ipck.msgp);
 		msgtyp = ipck.msgtyp;
 	}
-	p = compat_alloc_user_space(second + sizeof(struct msgbuf));
-	err = sys_msgrcv(first, p, second, msgtyp, third);
+	up = uptr;
+	err = do_msgrcv(first, &type, up->mtext, second, msgtyp, third);
 	if (err < 0)
 		goto out;
-	up = uptr;
-	if (get_user(type, &p->mtype) ||
-	    put_user(type, &up->mtype) ||
-	    copy_in_user(up->mtext, p->mtext, err))
+	if (put_user(type, &up->mtype))
 		err = -EFAULT;
 out:
 	return err;
ipc/msg.c +33 -11

86 lines not on this page · raw message