From: Alexander Viro <viro@math.psu.edu>
To: "Stephen C. Tweedie" <sct@redhat.com>
Cc: Manfred Spraul <manfreds@colorfullife.com>,
Andrea Arcangeli <andrea@suse.de>,
linux-kernel@vger.rutgers.edu,
Ingo Molnar <mingo@chiara.csoma.elte.hu>,
Linus Torvalds <torvalds@transmeta.com>,
linux-mm@kvack.org
Subject: [more fun] Re: locking question: do_mmap(), do_munmap()
Date: Tue, 12 Oct 1999 10:06:06 -0400 (EDT) [thread overview]
Message-ID: <Pine.GSO.4.10.9910120716490.22333-100000@weyl.math.psu.edu> (raw)
In-Reply-To: <Pine.GSO.4.10.9910111850370.18777-100000@weyl.math.psu.edu>
Funny path #1: on 386 (sucky WP) copy_to_user() -> access_ok() ->
__verify_write() -> handle_mm_fault() and no mmap_sem in sight. Ditto for
__verify_write() on sh.
Another one: ptrace_readdata() -> access_process_vm() -> find_extend_vma()
-> make_pages_present(). Again, no mmap_sem in sight.
irix_brk(): calls do_brk() without mmap_sem.
sys_cacheflush() (on m68k): plays with vma without mmap_sem.
Patch follows (2.3.20, but these files didn't change in .21).
diff -urN linux-2.3.20/arch/i386/mm/fault.c linux-bird.mm/arch/i386/mm/fault.c
--- linux-2.3.20/arch/i386/mm/fault.c Sun Sep 12 11:01:01 1999
+++ linux-bird.mm/arch/i386/mm/fault.c Tue Oct 12 07:44:55 1999
@@ -35,6 +35,7 @@
if (!size)
return 1;
+ down(¤t->mm->mmap_sem);
vma = find_vma(current->mm, start);
if (!vma)
goto bad_area;
@@ -64,6 +65,7 @@
if (!(vma->vm_flags & VM_WRITE))
goto bad_area;;
}
+ up(¤t->mm->mmap_sem);
return 1;
check_stack:
@@ -73,6 +75,7 @@
goto good_area;
bad_area:
+ up(¤t->mm->mmap_sem);
return 0;
}
diff -urN linux-2.3.20/arch/m68k/kernel/sys_m68k.c linux-bird.mm/arch/m68k/kernel/sys_m68k.c
--- linux-2.3.20/arch/m68k/kernel/sys_m68k.c Mon Jun 21 12:35:55 1999
+++ linux-bird.mm/arch/m68k/kernel/sys_m68k.c Tue Oct 12 09:56:24 1999
@@ -535,6 +535,7 @@
int ret = -EINVAL;
lock_kernel();
+ down(¤t->mm->mmap_sem);
if (scope < FLUSH_SCOPE_LINE || scope > FLUSH_SCOPE_ALL ||
cache & ~FLUSH_CACHE_BOTH)
goto out;
@@ -591,6 +592,7 @@
ret = cache_flush_060 (addr, scope, cache, len);
}
out:
+ up(¤t->mm->mmap_sem);
unlock_kernel();
return ret;
}
diff -urN linux-2.3.20/arch/mips/kernel/sysirix.c linux-bird.mm/arch/mips/kernel/sysirix.c
--- linux-2.3.20/arch/mips/kernel/sysirix.c Sun Sep 12 05:54:08 1999
+++ linux-bird.mm/arch/mips/kernel/sysirix.c Tue Oct 12 09:46:09 1999
@@ -534,6 +534,7 @@
int ret;
lock_kernel();
+ down(¤t->mm->mmap_sem);
if (brk < current->mm->end_code) {
ret = -ENOMEM;
goto out;
@@ -591,6 +592,7 @@
ret = 0;
out:
+ up(¤t->mm->mmap_sem);
unlock_kernel();
return ret;
}
diff -urN linux-2.3.20/arch/sh/mm/fault.c linux-bird.mm/arch/sh/mm/fault.c
--- linux-2.3.20/arch/sh/mm/fault.c Sun Sep 12 13:29:49 1999
+++ linux-bird.mm/arch/sh/mm/fault.c Tue Oct 12 09:57:03 1999
@@ -38,6 +38,7 @@
if (!size)
return 1;
+ down(¤t->mm->mmap_sem);
vma = find_vma(current->mm, start);
if (!vma)
goto bad_area;
@@ -67,6 +68,7 @@
if (!(vma->vm_flags & VM_WRITE))
goto bad_area;;
}
+ up(¤t->mm->mmap_sem);
return 1;
check_stack:
@@ -76,6 +78,7 @@
goto good_area;
bad_area:
+ up(¤t->mm->mmap_sem);
return 0;
}
diff -urN linux-2.3.20/kernel/ptrace.c linux-bird.mm/kernel/ptrace.c
--- linux-2.3.20/kernel/ptrace.c Sun Sep 12 13:03:24 1999
+++ linux-bird.mm/kernel/ptrace.c Tue Oct 12 09:15:27 1999
@@ -79,14 +79,15 @@
int access_process_vm(struct task_struct *tsk, unsigned long addr, void *buf, int len, int write)
{
- int copied;
- struct vm_area_struct * vma = find_extend_vma(tsk, addr);
+ int copied = 0;
+ struct vm_area_struct * vma;
+
+ down(&tsk->mm->mmap_sem);
+ vma = find_extend_vma(tsk, addr);
if (!vma)
- return 0;
+ goto out;
- down(&tsk->mm->mmap_sem);
- copied = 0;
for (;;) {
unsigned long offset = addr & ~PAGE_MASK;
int this_len = PAGE_SIZE - offset;
@@ -115,6 +116,7 @@
vma = vma->vm_next;
}
+out:
up(&tsk->mm->mmap_sem);
return copied;
}
--
To unsubscribe, send a message with 'unsubscribe linux-mm' in
the body to majordomo@kvack.org. For more info on Linux MM,
see: http://humbolt.geo.uu.nl/Linux-MM/
next prev parent reply other threads:[~1999-10-12 14:06 UTC|newest]
Thread overview: 43+ messages / expand[flat|nested] mbox.gz Atom feed top
[not found] <Pine.LNX.4.10.9910101713010.364-100000@alpha.random>
1999-10-10 15:52 ` Manfred Spraul
1999-10-10 16:07 ` Alexander Viro
1999-10-10 16:25 ` Alexander Viro
1999-10-10 16:45 ` Manfred Spraul
1999-10-10 17:25 ` Alexander Viro
1999-10-10 17:12 ` Andrea Arcangeli
1999-10-10 17:48 ` Alexander Viro
1999-10-10 18:42 ` Manfred Spraul
1999-10-10 19:03 ` Alexander Viro
1999-10-10 21:31 ` Manfred Spraul
1999-10-10 21:53 ` Andrea Arcangeli
1999-10-10 22:34 ` Alexander Viro
1999-10-10 23:28 ` Andrea Arcangeli
1999-10-11 15:50 ` Stephen C. Tweedie
1999-10-11 16:05 ` Alexander Viro
1999-10-11 18:02 ` Manfred Spraul
1999-10-11 19:07 ` Kanoj Sarcar
1999-10-11 22:23 ` Stephen C. Tweedie
1999-10-13 1:25 ` Kanoj Sarcar
1999-10-13 7:32 ` Manfred Spraul
1999-10-15 9:58 ` Ralf Baechle
1999-10-15 17:50 ` Kanoj Sarcar
1999-10-13 10:45 ` Stephen C. Tweedie
1999-10-11 20:15 ` Stephen C. Tweedie
1999-10-11 21:14 ` Manfred Spraul
1999-10-11 21:37 ` Alexander Viro
1999-10-11 22:13 ` Manfred Spraul
1999-10-11 22:22 ` Stephen C. Tweedie
1999-10-11 23:01 ` Alexander Viro
1999-10-12 14:06 ` Alexander Viro [this message]
1999-10-13 7:35 ` [more fun] " Manfred Spraul
1999-10-13 18:34 ` Kanoj Sarcar
1999-10-13 10:16 ` Stephen C. Tweedie
1999-10-11 20:13 ` Stephen C. Tweedie
1999-10-11 21:40 ` Alexander Viro
1999-10-11 22:20 ` Stephen C. Tweedie
1999-10-11 22:31 ` Alexander Viro
1999-10-13 10:25 ` Stephen C. Tweedie
1999-10-11 15:47 ` Stephen C. Tweedie
1999-10-11 15:43 ` Stephen C. Tweedie
1999-10-10 16:56 ` Andrea Arcangeli
1999-10-11 15:41 ` Stephen C. Tweedie
1999-10-11 15:52 ` Alexander Viro
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=Pine.GSO.4.10.9910120716490.22333-100000@weyl.math.psu.edu \
--to=viro@math.psu.edu \
--cc=andrea@suse.de \
--cc=linux-kernel@vger.rutgers.edu \
--cc=linux-mm@kvack.org \
--cc=manfreds@colorfullife.com \
--cc=mingo@chiara.csoma.elte.hu \
--cc=sct@redhat.com \
--cc=torvalds@transmeta.com \
/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