kernel.git/fs/proc/internal.h, branch linux-2.6.25.y

[NET]: Make /proc/net a symlink on /proc/self/net (v3)

2008-03-07T19:08:40Z

Current /proc/net is done with so called "shadows", but current implementation is broken and has little chances to get fixed. The problem is that dentries subtree of /proc/net directory has fancy revalidation rules to make processes living in different net namespaces see different entries in /proc/net subtree, but currently, tasks see in the /proc/net subdir the contents of any other namespace, depending on who opened the file first. The proposed fix is to turn /proc/net into a symlink, which points to /proc/self/net, which in turn shows what previously was in /proc/net - the network-related info, from the net namespace the appropriate task lives in. # ls -l /proc/net lrwxrwxrwx 1 root root 8 Mar 5 15:17 /proc/net -> self/net In other words - this behaves like /proc/mounts, but unlike "mounts", "net" is not a file, but a directory. Changes from v2: * Fixed discrepancy of /proc/net nlink count and selinux labeling screwup pointed out by Stephen. To get the correct nlink count the ->getattr callback for /proc/net is overridden to read one from the net->proc_net entry. To make selinux still work the net->proc_net entry is initialized properly, i.e. with the "net" name and the proc_net parent. Selinux fixes are Acked-by: Stephen Smalley Changes from v1: * Fixed a task_struct leak in get_proc_task_net, pointed out by Paul. Signed-off-by: Pavel Emelyanov Acked-by: "Eric W. Biederman" Signed-off-by: David S. Miller

d_path: Make proc_get_link() use a struct path argument

2008-02-15T05:17:08Z

proc_get_link() is always called with a dentry and a vfsmount from a struct path. Make proc_get_link() take it directly as an argument. Signed-off-by: Jan Blunck Acked-by: Christoph Hellwig Cc: Al Viro Cc: "J. Bruce Fields" Cc: Neil Brown Cc: "Eric W. Biederman" Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds

proc: seqfile convert proc_pid_status to properly handle pid namespaces

2008-02-08T17:22:24Z

Currently we possibly lookup the pid in the wrong pid namespace. So seq_file convert proc_pid_status which ensures the proper pid namespaces is passed in. [akpm@linux-foundation.org: coding-style fixes] [akpm@linux-foundation.org: build fix] [akpm@linux-foundation.org: another build fix] [akpm@linux-foundation.org: s390 build fix] [akpm@linux-foundation.org: fix task_name() output] [akpm@linux-foundation.org: fix nommu build] Signed-off-by: Eric W. Biederman Cc: Andrew Morgan Cc: Serge Hallyn Cc: Cedric Le Goater Cc: Pavel Emelyanov Cc: Martin Schwidefsky Cc: Heiko Carstens Cc: Paul Menage Cc: Paul Jackson Cc: David Rientjes Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds

seqfile convert proc_pid_statm

2008-02-08T17:22:24Z

This conversion is just for code cleanliness, uniformity, and general safety. Signed-off-by: Eric W. Biederman Cc: Oleg Nesterov Cc: Alexey Dobriyan Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds

proc: rewrite do_task_stat to correctly handle pid namespaces.

2008-02-08T17:22:23Z

Currently (as pointed out by Oleg) do_task_stat has a race when calling task_pid_nr_ns with the task exiting. In addition do_task_stat is not currently displaying information in the context of the pid namespace that mounted the /proc filesystem. So "cut -d' ' -f 1 /proc//stat" may not equal . This patch fixes the problem by converting to a single_open seq_file show method. Getting the pid namespace from the filesystem superblock instead of current, and simply using the the struct pid from the inode instead of attempting to get that same pid from the task. Signed-off-by: Eric W. Biederman Cc: Oleg Nesterov Cc: Alexey Dobriyan Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds

maps4: add /proc/pid/pagemap interface

2008-02-05T17:44:16Z

This interface provides a mapping for each page in an address space to its physical page frame number, allowing precise determination of what pages are mapped and what pages are shared between processes. New in this version: - headers gone again (as recommended by Dave Hansen and Alan Cox) - 64-bit entries (as per discussion with Andi Kleen) - swap pte information exported (from Dave Hansen) - page walker callback for holes (from Dave Hansen) - direct put_user I/O (as suggested by Rusty Russell) This patch folds in cleanups and swap PTE support from Dave Hansen . Signed-off-by: Matt Mackall Cc: Dave Hansen Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds

maps4: move clear_refs code to task_mmu.c

2008-02-05T17:44:16Z

This puts all the clear_refs code where it belongs and probably lets things compile on MMU-less systems as well. Signed-off-by: Matt Mackall Cc: Jeremy Fitzhardinge Cc: David Rientjes Cc: Dave Hansen Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds

restrict reading from /proc//maps to those who share ->mm or can ptrace pid

2008-01-02T21:13:27Z

Contents of /proc/*/maps is sensitive and may become sensitive after open() (e.g. if target originally shares our ->mm and later does exec on suid-root binary). Check at read() (actually, ->start() of iterator) time that mm_struct we'd grabbed and locked is - still the ->mm of target - equal to reader's ->mm or the target is ptracable by reader. Signed-off-by: Al Viro Acked-by: Rik van Riel Signed-off-by: Linus Torvalds

proc: fix NULL ->i_fop oops

2007-11-29T17:24:52Z

proc_kill_inodes() can clear ->i_fop in the middle of vfs_readdir resulting in NULL dereference during "file->f_op->readdir(file, buf, filler)". The solution is to remove proc_kill_inodes() completely: a) we don't have tricky modules implementing their tricky readdir hooks which could keeping this revoke from hell. b) In a situation when module is gone but PDE still alive, standard readdir will return only "." and "..", because pde->next was cleared by remove_proc_entry(). c) the race proc_kill_inode() destined to prevent is not completely fixed, just race window made smaller, because vfs_readdir() is run without sb_lock held and without file_list_lock held. Effectively, ->i_fop is cleared at random moment, which can't fix properly anything. BUG: unable to handle kernel NULL pointer dereference at virtual address 00000018 printing eip: c1061205 *pdpt = 0000000005b22001 *pde = 0000000000000000 Oops: 0000 [#1] PREEMPT SMP Modules linked in: foo af_packet ipv6 cpufreq_ondemand loop serio_raw sr_mod k8temp cdrom hwmon amd_rng Pid: 2033, comm: find Not tainted (2.6.24-rc1-b1d08ac064268d0ae2281e98bf5e82627e0f0c56 #2) EIP: 0060:[] EFLAGS: 00010246 CPU: 0 EIP is at vfs_readdir+0x47/0x74 EAX: c6b6a780 EBX: 00000000 ECX: c1061040 EDX: c5decf94 ESI: c6b6a780 EDI: fffffffe EBP: c9797c54 ESP: c5decf78 DS: 007b ES: 007b FS: 00d8 GS: 0033 SS: 0068 Process find (pid: 2033, ti=c5dec000 task=c64bba90 task.ti=c5dec000) Stack: c5decf94 c1061040 fffffff7 0805ffbc 00000000 c6b6a780 c1061295 0805ffbc 00000000 00000400 00000000 00000004 0805ffbc 4588eff4 c5dec000 c10026ba 00000004 0805ffbc 00000400 0805ffbc 4588eff4 bfdc6c70 000000dc 0000007b Call Trace: [] filldir64+0x0/0xc5 [] sys_getdents64+0x63/0xa5 [] sysenter_past_esp+0x5f/0x85 ======================= Code: 49 83 78 18 00 74 43 8d 6b 74 bf fe ff ff ff 89 e8 e8 b8 c0 12 00 f6 83 2c 01 00 00 10 75 22 8b 5e 10 8b 4c 24 04 89 f0 8b 14 24 53 18 f6 46 1a 04 89 c7 75 0b 8b 56 0c 8b 46 08 e8 c8 66 00 EIP: [] vfs_readdir+0x47/0x74 SS:ESP 0068:c5decf78 hch: "Nice, getting rid of this is a very good step formwards. Unfortunately we have another copy of this junk in security/selinux/selinuxfs.c:sel_remove_entries() which would need the same treatment." Signed-off-by: Alexey Dobriyan Acked-by: Christoph Hellwig Cc: Al Viro Cc: Stephen Smalley Cc: James Morris Cc: "Eric W. Biederman" Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds

proc: fix proc_kill_inodes to kill dentries on all proc superblocks

2007-11-15T02:45:38Z

It appears we overlooked support for removing generic proc files when we added support for multiple proc super blocks. Handle that now. [akpm@linux-foundation.org: coding-style cleanups] Signed-off-by: Eric W. Biederman Acked-by: Pavel Emelyanov Cc: Alexey Dobriyan Acked-by: Sukadev Bhattiprolu Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds