[lkml]   [2017]   [Dec]   [13]   [last100]   RSS Feed
Views: [wrap][no wrap]   [headers]  [forward] 
Messages in this thread
SubjectRe: [PATCH 2/2] mmap.2: MAP_FIXED updated documentation
On Wed, Dec 13, 2017 at 10:31 AM, Michal Hocko <> wrote:
> From: John Hubbard <>
> -- Expand the documentation to discuss the hazards in
> enough detail to allow avoiding them.
> -- Mention the upcoming MAP_FIXED_SAFE flag.
> -- Enhance the alignment requirement slightly.
> CC: Michael Ellerman <>
> CC: Jann Horn <>
> CC: Matthew Wilcox <>
> CC: Michal Hocko <>
> CC: Mike Rapoport <>
> CC: Cyril Hrubis <>
> CC: Pavel Machek <>
> Acked-by: Michal Hocko <>
> Signed-off-by: John Hubbard <>
> Signed-off-by: Michal Hocko <>
> ---
> man2/mmap.2 | 32 ++++++++++++++++++++++++++++++--
> 1 file changed, 30 insertions(+), 2 deletions(-)
> diff --git a/man2/mmap.2 b/man2/mmap.2
> index 02d391697ce6..cb8789daec2d 100644
> --- a/man2/mmap.2
> +++ b/man2/mmap.2
> @@ -226,6 +227,33 @@ Software that aspires to be portable should use this option with care, keeping
> in mind that the exact layout of a process' memory map is allowed to change
> significantly between kernel versions, C library versions, and operating system
> releases.
> +.IP
> +Furthermore, this option is extremely hazardous (when used on its own), because
> +it forcibly removes pre-existing mappings, making it easy for a multi-threaded
> +process to corrupt its own address space.

I think this is worded unfortunately. It is dangerous if used
incorrectly, and it's a good tool when used correctly.

> +Thread B need not create a mapping directly; simply making a library call
> +that, internally, uses
> +.I dlopen(3)
> +to load some other shared library, will
> +suffice. The dlopen(3) call will map the library into the process's address
> +space. Furthermore, almost any library call may be implemented using this
> +technique.
> +Examples include brk(2), malloc(3), pthread_create(3), and the PAM libraries
> +(

This is arkward. This first mentions dlopen(), which is a very niche
case, and then just very casually mentions the much bigger
problem that tons of library functions can allocate memory through
malloc(), causing mmap() calls, sometimes without that even being
a documented property of the function.

> +.IP
> +Newer kernels
> +(Linux 4.16 and later) have a
> +option that avoids the corruption problem; if available, MAP_FIXED_SAFE
> +should be preferred over MAP_FIXED.

This is bad advice. MAP_FIXED is completely safe if you use it on an address
range you've allocated, and it is used in this way by core system libraries to
place multiple VMAs in virtually contiguous memory, for example: (from glibc) uses it to load dynamic libraries:

$ strace -e trace=open,mmap,close /usr/bin/id 2>&1 >/dev/null | head -n20
0) = 0x7f35811c0000
open("/etc/", O_RDONLY|O_CLOEXEC) = 3
mmap(NULL, 161237, PROT_READ, MAP_PRIVATE, 3, 0) = 0x7f3581198000
close(3) = 0
open("/lib/x86_64-linux-gnu/", O_RDONLY|O_CLOEXEC) = 3
0) = 0x7f3580d78000
mmap(0x7f3580f9c000, 8192, PROT_READ|PROT_WRITE,
MAP_PRIVATE|MAP_FIXED|MAP_DENYWRITE, 3, 0x24000) = 0x7f3580f9c000
mmap(0x7f3580f9e000, 6864, PROT_READ|PROT_WRITE,
MAP_PRIVATE|MAP_FIXED|MAP_ANONYMOUS, -1, 0) = 0x7f3580f9e000
close(3) = 0
open("/lib/x86_64-linux-gnu/", O_RDONLY|O_CLOEXEC) = 3
0) = 0x7f35809d9000
mmap(0x7f3580d6e000, 24576, PROT_READ|PROT_WRITE,
MAP_PRIVATE|MAP_FIXED|MAP_DENYWRITE, 3, 0x195000) = 0x7f3580d6e000
mmap(0x7f3580d74000, 14752, PROT_READ|PROT_WRITE,
MAP_PRIVATE|MAP_FIXED|MAP_ANONYMOUS, -1, 0) = 0x7f3580d74000
close(3) = 0

As a comment in dl-map-segments.h in glibc explains:
/* This is a position-independent shared object. We can let the
kernel map it anywhere it likes, but we must have space for all
the segments in their specified positions relative to the first.
So we map the first segment without MAP_FIXED, but with its
extent increased to cover all the segments. Then we remove
access from excess portion, and there is known sufficient space
there to remap from the later segments.

And AFAIK anything that allocates thread stacks uses MAP_FIXED to
create the guard page at the bottom.

MAP_FIXED is a better solution for these usecases than MAP_FIXED_SAFE,
or whatever it ends up being called. Please remove this advice or, better,
clarify what MAP_FIXED should be used for (creation of virtually contiguous
VMAs) and what MAP_FIXED_SAFE should be used for (attempting to
allocate memory at a fixed address for some reason, with a failure instead of
the normal fallback to using a different address).

 \ /
  Last update: 2017-12-14 03:54    [W:0.143 / U:0.000 seconds]
©2003-2020 Jasper Spaans|hosted at Digital Ocean and TransIP|Read the blog|Advertise on this site