Make sure all dentries refs are released before calling kill_anon_super()
so that the assumption that generic_shutdown_super() can completely destroy
the dentry tree for there will be no external references holds true.
What was being done in the put_super() superblock op, is now done in the
kill_sb() filesystem op instead, prior to calling kill_anon_super().
The call to shrink_dcache_sb() is removed as it is redundant since
shrink_dcache_for_umount() will now be called after the cleanup routine.
Signed-off-by: David Howells <[email protected]>
Acked-by: Ian Kent <[email protected]>
Cc: Andi Kleen <[email protected]>
Signed-off-by: Andrew Morton <[email protected]>
Signed-off-by: Linus Torvalds <[email protected]>
Use inc/dec_preempt_count() rather than preempt_enable/disable() and manually
add in the compiler barriers that were provided by the latter. This makes FRV
consistent with other archs.
Furthermore, the compiler barrier effects are now there unconditionally - at
least as far as preemption is concerned - because we don't want the compiler
moving memory accesses out of the section of code in which the mapping is in
force - in effect the kmap_atomic() must imply a LOCK-class barrier and the
kunmap_atomic() must imply an UNLOCK-class barrier to the compiler.
Signed-off-by: David Howells <[email protected]>
Acked-by: Peter Zijlstra <[email protected]>
Signed-off-by: Linus Torvalds <[email protected]>
The attached patch destroys all the dentries attached to a superblock in one go
by:
(1) Destroying the tree rooted at s_root.
(2) Destroying every entry in the anon list, one at a time.
(3) Each entry in the anon list has its subtree consumed from the leaves
inwards.
This reduces the amount of work generic_shutdown_super() does, and avoids
iterating through the dentry_unused list.
Note that locking is almost entirely absent in the shrink_dcache_for_umount*()
functions added by this patch. This is because:
(1) at the point the filesystem calls generic_shutdown_super(), it is not
permitted to further touch the superblock's set of dentries, and nor may
it remove aliases from inodes;
(2) the dcache memory shrinker now skips dentries that are being unmounted;
and
(3) the superblock no longer has any external references through which the VFS
can reach it.
Given these points, the only locking we need to do is when we remove dentries
from the unused list and the name hashes, which we do a directory's worth at a
time.
We also don't need to guard against reference counts going to zero unexpectedly
and removing bits of the tree we're working on as nothing else can call dput().
A cut down version of dentry_iput() has been folded into
shrink_dcache_for_umount_subtree() function. Apart from not needing to unlock
things, it also doesn't need to check for inotify watches.
In this version of the patch, the complaint about a dentry still being in use
has been expanded from a single BUG_ON() and now gives much more information.
Signed-off-by: David Howells <[email protected]>
Acked-by: NeilBrown <[email protected]>
Acked-by: Ian Kent <[email protected]>
Cc: Trond Myklebust <[email protected]>
Signed-off-by: Andrew Morton <[email protected]>
Signed-off-by: Linus Torvalds <[email protected]>
Make sure all dentries refs are released before calling kill_anon_super() so
that the assumption that generic_shutdown_super() can completely destroy the
dentry tree for there will be no external references holds true.
What was being done in the put_super() superblock op, is now done in the
kill_sb() filesystem op instead, prior to calling kill_anon_super().
This makes the struct autofs_sb_info::root member variable redundant (since
sb->s_root is still available), and so that is removed. The calls to
shrink_dcache_sb() are also removed since they're also redundant as
shrink_dcache_for_umount() will now be called after the cleanup routine.
Signed-off-by: David Howells <[email protected]>
Acked-by: Ian Kent <[email protected]>
Cc: Trond Myklebust <[email protected]>
Signed-off-by: Andrew Morton <[email protected]>
Signed-off-by: Linus Torvalds <[email protected]>
Make sure all dentries refs are released before calling kill_block_super()
so that the assumption that generic_shutdown_super() can completely destroy
the dentry tree for there will be no external references holds true.
What was being done in the put_super() superblock op, is now done in the
kill_sb() filesystem op instead, prior to calling kill_block_super().
Changes made in [try #2]:
(*) reiserfs_kill_sb() now checks that the superblock FS info pointer is set
before trying to dereference it.
Signed-off-by: David Howells <[email protected]>
Cc: "Rafael J. Wysocki" <[email protected]>
Cc: <[email protected]>
Signed-off-by: Andrew Morton <[email protected]>
Signed-off-by: Linus Torvalds <[email protected]>
Use the new typedef for interrupt handler function pointers rather than
actually spelling out the full thing each time. This was scripted with the
following small shell script:
#!/bin/sh
egrep -nHrl -e 'irqreturn_t[ ]*[(][*]' $* |
while read i
do
echo $i
perl -pi -e 's/irqreturn_t\s*[(]\s*[*]\s*([_a-zA-Z0-9]*)\s*[)]\s*[(]\s*int\s*,\s*void\s*[*]\s*[)]/irq_handler_t \1/g' $i || exit $?
done
Signed-Off-By: David Howells <[email protected]>
Maintain a per-CPU global "struct pt_regs *" variable which can be used instead
of passing regs around manually through all ~1800 interrupt handlers in the
Linux kernel.
The regs pointer is used in few places, but it potentially costs both stack
space and code to pass it around. On the FRV arch, removing the regs parameter
from all the genirq function results in a 20% speed up of the IRQ exit path
(ie: from leaving timer_interrupt() to leaving do_IRQ()).
Where appropriate, an arch may override the generic storage facility and do
something different with the variable. On FRV, for instance, the address is
maintained in GR28 at all times inside the kernel as part of general exception
handling.
Having looked over the code, it appears that the parameter may be handed down
through up to twenty or so layers of functions. Consider a USB character
device attached to a USB hub, attached to a USB controller that posts its
interrupts through a cascaded auxiliary interrupt controller. A character
device driver may want to pass regs to the sysrq handler through the input
layer which adds another few layers of parameter passing.
I've build this code with allyesconfig for x86_64 and i386. I've runtested the
main part of the code on FRV and i386, though I can't test most of the drivers.
I've also done partial conversion for powerpc and MIPS - these at least compile
with minimal configurations.
This will affect all archs. Mostly the changes should be relatively easy.
Take do_IRQ(), store the regs pointer at the beginning, saving the old one:
struct pt_regs *old_regs = set_irq_regs(regs);
And put the old one back at the end:
set_irq_regs(old_regs);
Don't pass regs through to generic_handle_irq() or __do_IRQ().
In timer_interrupt(), this sort of change will be necessary:
- update_process_times(user_mode(regs));
- profile_tick(CPU_PROFILING, regs);
+ update_process_times(user_mode(get_irq_regs()));
+ profile_tick(CPU_PROFILING);
I'd like to move update_process_times()'s use of get_irq_regs() into itself,
except that i386, alone of the archs, uses something other than user_mode().
Some notes on the interrupt handling in the drivers:
(*) input_dev() is now gone entirely. The regs pointer is no longer stored in
the input_dev struct.
(*) finish_unlinks() in drivers/usb/host/ohci-q.c needs checking. It does
something different depending on whether it's been supplied with a regs
pointer or not.
(*) Various IRQ handler function pointers have been moved to type
irq_handler_t.
Signed-Off-By: David Howells <[email protected]>
(cherry picked from 1b16e7ac850969f38b375e511e3fa2f474a33867 commit)
Typedef the IRQ handler function type.
Signed-Off-By: David Howells <[email protected]>
(cherry picked from 1356d1e5fd256997e3d3dce0777ab787d0515c7a commit)
Typedef the IRQ flow handler function type.
Signed-Off-By: David Howells <[email protected]>
(cherry picked from 8e973fbdf5716b93a0a8c0365be33a31ca0fa351 commit)
Permit kmalloc() to make allocations of up to 32MB if so configured. This
may be useful under NOMMU conditions where vmalloc() can't do this.
Signed-off-by: David Howells <[email protected]>
Signed-off-by: Andrew Morton <[email protected]>
Signed-off-by: Linus Torvalds <[email protected]>
These patches make the kernel pass 64-bit inode numbers internally when
communicating to userspace, even on a 32-bit system. They are required
because some filesystems have intrinsic 64-bit inode numbers: NFS3+ and XFS
for example. The 64-bit inode numbers are then propagated to userspace
automatically where the arch supports it.
Problems have been seen with userspace (eg: ld.so) using the 64-bit inode
number returned by stat64() or getdents64() to differentiate files, and
failing because the 64-bit inode number space was compressed to 32-bits, and
so overlaps occur.
This patch:
Make filldir_t take a 64-bit inode number and struct kstat carry a 64-bit
inode number so that 64-bit inode numbers can be passed back to userspace.
The stat functions then returns the full 64-bit inode number where
available and where possible. If it is not possible to represent the inode
number supplied by the filesystem in the field provided by userspace, then
error EOVERFLOW will be issued.
Similarly, the getdents/readdir functions now pass the full 64-bit inode
number to userspace where possible, returning EOVERFLOW instead when a
directory entry is encountered that can't be properly represented.
Note that this means that some inodes will not be stat'able on a 32-bit
system with old libraries where they were before - but it does mean that
there will be no ambiguity over what a 32-bit inode number refers to.
Note similarly that directory scans may be cut short with an error on a
32-bit system with old libraries where the scan would work before for the
same reasons.
It is judged unlikely that this situation will occur because modern glibc
uses 64-bit capable versions of stat and getdents class functions
exclusively, and that older systems are unlikely to encounter
unrepresentable inode numbers anyway.
[akpm: alpha build fix]
Signed-off-by: David Howells <[email protected]>
Cc: Trond Myklebust <[email protected]>
Cc: Al Viro <[email protected]>
Signed-off-by: Andrew Morton <[email protected]>
Signed-off-by: Linus Torvalds <[email protected]>
Revert Andrew Morton's patch to temporarily hack around the lack of a
declaration of sigset_t in linux/compat.h to make the block-disablement
patches build on IA64. This got accidentally pushed to Linus and should
be fixed in a different manner.
Also make linux/compat.h #include asm/signal.h to gain a definition of
sigset_t so that it can externally declare sigset_from_compat().
This has been compile-tested for i386, x86_64, ia64, mips, mips64, frv, ppc and
ppc64 and run-tested on frv.
Signed-off-by: David Howells <[email protected]>
Signed-off-by: Linus Torvalds <[email protected]>
This makes CONFIG_USB_STORAGE depend on CONFIG_SCSI rather than selecting it,
as selecting it makes CONFIG_USB_STORAGE override the dependencies of SCSI,
causing it to turn on even if they aren't all met.
Signed-Off-By: David Howells <[email protected]>
Signed-off-by: Jens Axboe <[email protected]>
Make it possible to disable the block layer. Not all embedded devices require
it, some can make do with just JFFS2, NFS, ramfs, etc - none of which require
the block layer to be present.
This patch does the following:
(*) Introduces CONFIG_BLOCK to disable the block layer, buffering and blockdev
support.
(*) Adds dependencies on CONFIG_BLOCK to any configuration item that controls
an item that uses the block layer. This includes:
(*) Block I/O tracing.
(*) Disk partition code.
(*) All filesystems that are block based, eg: Ext3, ReiserFS, ISOFS.
(*) The SCSI layer. As far as I can tell, even SCSI chardevs use the
block layer to do scheduling. Some drivers that use SCSI facilities -
such as USB storage - end up disabled indirectly from this.
(*) Various block-based device drivers, such as IDE and the old CDROM
drivers.
(*) MTD blockdev handling and FTL.
(*) JFFS - which uses set_bdev_super(), something it could avoid doing by
taking a leaf out of JFFS2's book.
(*) Makes most of the contents of linux/blkdev.h, linux/buffer_head.h and
linux/elevator.h contingent on CONFIG_BLOCK being set. sector_div() is,
however, still used in places, and so is still available.
(*) Also made contingent are the contents of linux/mpage.h, linux/genhd.h and
parts of linux/fs.h.
(*) Makes a number of files in fs/ contingent on CONFIG_BLOCK.
(*) Makes mm/bounce.c (bounce buffering) contingent on CONFIG_BLOCK.
(*) set_page_dirty() doesn't call __set_page_dirty_buffers() if CONFIG_BLOCK
is not enabled.
(*) fs/no-block.c is created to hold out-of-line stubs and things that are
required when CONFIG_BLOCK is not set:
(*) Default blockdev file operations (to give error ENODEV on opening).
(*) Makes some /proc changes:
(*) /proc/devices does not list any blockdevs.
(*) /proc/diskstats and /proc/partitions are contingent on CONFIG_BLOCK.
(*) Makes some compat ioctl handling contingent on CONFIG_BLOCK.
(*) If CONFIG_BLOCK is not defined, makes sys_quotactl() return -ENODEV if
given command other than Q_SYNC or if a special device is specified.
(*) In init/do_mounts.c, no reference is made to the blockdev routines if
CONFIG_BLOCK is not defined. This does not prohibit NFS roots or JFFS2.
(*) The bdflush, ioprio_set and ioprio_get syscalls can now be absent (return
error ENOSYS by way of cond_syscall if so).
(*) The seclvl_bd_claim() and seclvl_bd_release() security calls do nothing if
CONFIG_BLOCK is not set, since they can't then happen.
Signed-Off-By: David Howells <[email protected]>
Signed-off-by: Jens Axboe <[email protected]>
Remove inclusions of linux/buffer_head.h that are no longer necessary due to the
transfer of a number of things out of there.
Signed-Off-By: David Howells <[email protected]>
Signed-off-by: Jens Axboe <[email protected]>
Remove inclusions of linux/mpage.h that are no longer necessary due to the
transfer of generic_writepages().
Signed-Off-By: David Howells <[email protected]>
Signed-off-by: Jens Axboe <[email protected]>
Move the msdos device ioctl compat stuff from fs/compat_ioctl.c to the msdos
driver so that the msdos header file doesn't need to be included.
Signed-Off-By: David Howells <[email protected]>
Signed-off-by: Jens Axboe <[email protected]>
Move the Ext3 device ioctl compat stuff from fs/compat_ioctl.c to the Ext3
driver so that the Ext3 header file doesn't need to be included.
Signed-Off-By: David Howells <[email protected]>
Signed-off-by: Jens Axboe <[email protected]>
Move the Ext2 device ioctl compat stuff from fs/compat_ioctl.c to the Ext2
driver so that the Ext2 header file doesn't need to be included.
Signed-Off-By: David Howells <[email protected]>
Signed-off-by: Jens Axboe <[email protected]>
Move the ReiserFS device ioctl compat stuff from fs/compat_ioctl.c to the
ReiserFS driver so that the ReiserFS header file doesn't need to be included.
Signed-Off-By: David Howells <[email protected]>
Signed-off-by: Jens Axboe <[email protected]>
Move common FS-specific ioctls from linux/ext2_fs.h to linux/fs.h as FS_IOC_*
and FS_IOC32_* and have the users of them use those as a base.
Also move the GETFLAGS/SETFLAGS flags to linux/fs.h as FS_*_FL macros, and then
have the other users use them as a base.
Signed-Off-By: David Howells <[email protected]>
Signed-off-by: Jens Axboe <[email protected]>
Move the loop device ioctl compat stuff from fs/compat_ioctl.c to the loop
driver so that the loop header file doesn't need to be included.
Signed-Off-By: David Howells <[email protected]>
Signed-off-by: Jens Axboe <[email protected]>
Move __invalidate_device() from fs/inode.c to fs/block_dev.c so that it can
more easily be disabled when the block layer is disabled.
Signed-Off-By: David Howells <[email protected]>
Signed-off-by: Jens Axboe <[email protected]>
Dissociate the generic_writepages() function from the mpage stuff, moving its
declaration to linux/mm.h and actually emitting a full implementation into
mm/page-writeback.c.
The implementation is a partial duplicate of mpage_writepages() with all BIO
references removed.
It is used by NFS to do writeback.
Signed-Off-By: David Howells <[email protected]>
Signed-off-by: Jens Axboe <[email protected]>
Move blockdev_superblock extern declaration from fs/fs-writeback.c to a
headerfile and remove the dependence on it by wrapping it in a macro.
Signed-Off-By: David Howells <[email protected]>
Signed-off-by: Jens Axboe <[email protected]>
Create a new header file, fs/internal.h, for common definitions local to the
sources in the fs/ directory.
Move extern definitions that should be in header files from fs/*.c to
fs/internal.h or other main header files where they span directories.
Signed-Off-By: David Howells <[email protected]>
Signed-off-by: Jens Axboe <[email protected]>
The AFS filesystem no longer needs to override its sync_page() op.
Signed-Off-By: David Howells <[email protected]>
Signed-off-by: Jens Axboe <[email protected]>
Move the bounce buffer code from mm/highmem.c to mm/bounce.c so that it can be
more easily disabled when the block layer is disabled.
!!!NOTE!!! There may be a bug in this code: Should init_emergency_pool() be
contingent on CONFIG_HIGHMEM?
Signed-Off-By: David Howells <[email protected]>
Signed-off-by: Jens Axboe <[email protected]>
Stop fallback_migrate_page() from using page_has_buffers() since that might not
be available. Use PagePrivate() instead since that's more general.
Signed-Off-By: David Howells <[email protected]>
Signed-off-by: Jens Axboe <[email protected]>
Remove the duplicate declaration of exit_io_context() from linux/sched.h.
Signed-Off-By: David Howells <[email protected]>
Signed-off-by: Jens Axboe <[email protected]>
Move some functions out of the buffering code that aren't strictly buffering
specific. This is a precursor to being able to disable the block layer.
(*) Moved some stuff out of fs/buffer.c:
(*) The file sync and general sync stuff moved to fs/sync.c.
(*) The superblock sync stuff moved to fs/super.c.
(*) do_invalidatepage() moved to mm/truncate.c.
(*) try_to_release_page() moved to mm/filemap.c.
(*) Moved some related declarations between header files:
(*) declarations for do_invalidatepage() and try_to_release_page() moved
to linux/mm.h.
(*) __set_page_dirty_buffers() moved to linux/buffer_head.h.
Signed-Off-By: David Howells <[email protected]>
Signed-off-by: Jens Axboe <[email protected]>
Move the fallback arch_vma_name() to a sensible place (kernel/signal.c).
Currently it's in fs/proc/task_mmu.c, a file that is dependent on both
CONFIG_PROC_FS and CONFIG_MMU being enabled, but it's used from
kernel/signal.c from where it is called unconditionally.
[[email protected]: build fix]
Signed-off-by: David Howells <[email protected]>
Signed-off-by: Andrew Morton <[email protected]>
Signed-off-by: Linus Torvalds <[email protected]>
Make futexes work under NOMMU conditions.
This can be tested by running this in one shell:
#define SYSERROR(X, Y) \
do { if ((long)(X) == -1L) { perror(Y); exit(1); }} while(0)
int main()
{
int shmid, tmp, *f, n;
shmid = shmget(23, 4, IPC_CREAT|0666);
SYSERROR(shmid, "shmget");
f = shmat(shmid, NULL, 0);
SYSERROR(f, "shmat");
n = *f;
printf("WAIT: %p{%x}\n", f, n);
tmp = futex(f, FUTEX_WAIT, n, NULL, NULL, 0);
SYSERROR(tmp, "futex");
printf("WAITED: %d\n", tmp);
tmp = shmdt(f);
SYSERROR(tmp, "shmdt");
exit(0);
}
And then this in the other shell:
#define SYSERROR(X, Y) \
do { if ((long)(X) == -1L) { perror(Y); exit(1); }} while(0)
int main()
{
int shmid, tmp, *f;
shmid = shmget(23, 4, IPC_CREAT|0666);
SYSERROR(shmid, "shmget");
f = shmat(shmid, NULL, 0);
SYSERROR(f, "shmat");
(*f)++;
printf("WAKE: %p{%x}\n", f, *f);
tmp = futex(f, FUTEX_WAKE, 1, NULL, NULL, 0);
SYSERROR(tmp, "futex");
printf("WOKE: %d\n", tmp);
tmp = shmdt(f);
SYSERROR(tmp, "shmdt");
exit(0);
}
The first program will set up a SYSV IPC SHM segment and wait on a futex in it
for the number at the start to change. The program will increment that number
and wake the first program up. This leads to output of the form:
SHELL 1 SHELL 2
======================= =======================
# /dowait
WAIT: 0xc32ac000{0}
# /dowake
WAKE: 0xc32ac000{1}
WAITED: 0 WOKE: 1
Signed-off-by: David Howells <[email protected]>
Signed-off-by: Andrew Morton <[email protected]>
Signed-off-by: Linus Torvalds <[email protected]>
Add documentation about using shared memory in NOMMU mode.
Signed-off-by: David Howells <[email protected]>
Signed-off-by: Andrew Morton <[email protected]>
Signed-off-by: Linus Torvalds <[email protected]>
Make mremap() partially work for NOMMU kernels. It may resize a VMA provided
that it doesn't exceed the size of the slab object in which the storage is
allocated that the VMA refers to. Shareable VMAs may not be resized.
Moving VMAs (as permitted by MREMAP_MAYMOVE) is not currently supported.
This patch also makes use of the fact that the VMA list is now ordered to cut
it short when possible.
Signed-off-by: David Howells <[email protected]>
Signed-off-by: Andrew Morton <[email protected]>
Signed-off-by: Linus Torvalds <[email protected]>
Order the per-mm_struct VMA list by address so that searching it can be cut
short when the appropriate address has been exceeded.
Signed-off-by: David Howells <[email protected]>
Signed-off-by: Andrew Morton <[email protected]>
Signed-off-by: Linus Torvalds <[email protected]>
Implement /proc/pid/maps for NOMMU by reading the vm_area_list attached to
current->mm->context.vmlist.
Signed-off-by: David Howells <[email protected]>
Signed-off-by: Andrew Morton <[email protected]>
Signed-off-by: Linus Torvalds <[email protected]>
Permit ptrace to modify a section that's non-shared but is marked
unwritable, such as is obtained by mapping the text segment of an ELF-FDPIC
executable binary with into a binary that's being ptraced[*].
[*] Under NOMMU conditions ptrace causes read-only MAP_PRIVATE mmaps to become
totally private copies because if a private mapping was actually shared
then the debugging setting breakpoints in it would potentially crash
other processes.
This is done by using the VM_MAYWRITE flag rather than the VM_WRITE flag
when deciding whether to permit a write.
Without this patch a debugger can't set breakpoints in the mapped text
sections of executables that are mapped read-only private, even if the
mmap() syscall has taken a private copy because PT_PTRACED is set.
In addition, VM_MAYREAD is used instead of VM_READ for similar reasons.
Signed-off-by: David Howells <[email protected]>
Signed-off-by: Andrew Morton <[email protected]>
Signed-off-by: Linus Torvalds <[email protected]>
Check the VMA protections in get_user_pages() against what's being asked.
This checks to see that we don't accidentally write on a non-writable VMA or
permit an I/O mapping VMA to be accessed (which may lack page structs).
Signed-off-by: David Howells <[email protected]>
Signed-off-by: Andrew Morton <[email protected]>
Signed-off-by: Linus Torvalds <[email protected]>
Use find_vma() in the NOMMU version of access_process_vm() rather than
reimplementing it.
Signed-off-by: David Howells <[email protected]>
Signed-off-by: Andrew Morton <[email protected]>
Signed-off-by: Linus Torvalds <[email protected]>
Set the backing device info capabilities for /dev/mem and /dev/kmem to
permit direct sharing under no-MMU conditions and full mapping capabilities
under MMU conditions. Make the BDI used by these available to all directly
mappable character devices.
Also comment the capabilities for /dev/zero.
[[email protected]: ifdef reductions]
Signed-off-by: David Howells <[email protected]>
Signed-off-by: Andrew Morton <[email protected]>
Signed-off-by: Linus Torvalds <[email protected]>
Check that access_process_vm() is accessing a valid mapping in the target
process.
This limits ptrace() accesses and accesses through /proc/<pid>/maps to only
those regions actually mapped by a program.
Signed-off-by: David Howells <[email protected]>
Signed-off-by: Andrew Morton <[email protected]>
Signed-off-by: Linus Torvalds <[email protected]>
Optimise ffs(x) by using fls(x & x - 1) which we optimise to use the SCAN
instruction.
Signed-off-by: David Howells <[email protected]>
Signed-off-by: Andrew Morton <[email protected]>
Signed-off-by: Linus Torvalds <[email protected]>
Implement fls64() for FRV without recource to conditional jumps.
Signed-off-by: David Howells <[email protected]>
Signed-off-by: Andrew Morton <[email protected]>
Signed-off-by: Linus Torvalds <[email protected]>
Fix FRV fls() to handle bit 31 being set correctly (it should return 32 not 0).
Signed-off-by: David Howells <[email protected]>
Signed-off-by: Andrew Morton <[email protected]>
Signed-off-by: Linus Torvalds <[email protected]>
Improve FRV's use of generic IRQ handling:
(*) Use generic_handle_irq() rather than __do_IRQ() as the latter is obsolete.
(*) Don't implement enable() and disable() ops as these will fall back to
using unmask() and mask().
(*) Provide mask_ack() functions to avoid a call each to mask() and ack().
(*) Make the cascade handlers always return IRQ_HANDLED.
(*) Implement the mask() and unmask() functions in the same order as they're
listed in the ops table.
Signed-off-by: David Howells <[email protected]>
Cc: Benjamin Herrenschmidt <[email protected]>
Cc: Thomas Gleixner <[email protected]>
Cc: Ingo Molnar <[email protected]>
Signed-off-by: Andrew Morton <[email protected]>
Signed-off-by: Linus Torvalds <[email protected]>
Make the FRV arch use the generic IRQ code rather than having its own
routines for doing so.
Signed-off-by: David Howells <[email protected]>
Signed-off-by: Andrew Morton <[email protected]>
Signed-off-by: Linus Torvalds <[email protected]>
Invoke security_d_instantiate() on root dentries after allocating them with
dentry_alloc_anon(). Normally dentry_alloc_root() would do that, but we don't
call that as we don't want to assign a name to the root dentry at this point
(we may discover the real name later).
Signed-Off-By: David Howells <[email protected]>
Signed-off-by: Trond Myklebust <[email protected]>
Fix an error handling problem: nfs_put_client() can be given a NULL pointer if
nfs_free_server() is asked to destroy a partially initialised record.
Signed-Off-By: David Howells <[email protected]>
Signed-off-by: Trond Myklebust <[email protected]>
Make two new proc files available:
/proc/fs/nfsfs/servers
/proc/fs/nfsfs/volumes
The first lists the servers with which we are currently dealing (struct
nfs_client), and the second lists the volumes we have on those servers (struct
nfs_server).
Signed-Off-By: David Howells <[email protected]>
Signed-off-by: Trond Myklebust <[email protected]>
The attached patch makes NFS share superblocks between mounts from the same
server and FSID over the same protocol.
It does this by creating each superblock with a false root and returning the
real root dentry in the vfsmount presented by get_sb(). The root dentry set
starts off as an anonymous dentry if we don't already have the dentry for its
inode, otherwise it simply returns the dentry we already have.
We may thus end up with several trees of dentries in the superblock, and if at
some later point one of anonymous tree roots is discovered by normal filesystem
activity to be located in another tree within the superblock, the anonymous
root is named and materialises attached to the second tree at the appropriate
point.
Why do it this way? Why not pass an extra argument to the mount() syscall to
indicate the subpath and then pathwalk from the server root to the desired
directory? You can't guarantee this will work for two reasons:
(1) The root and intervening nodes may not be accessible to the client.
With NFS2 and NFS3, for instance, mountd is called on the server to get
the filehandle for the tip of a path. mountd won't give us handles for
anything we don't have permission to access, and so we can't set up NFS
inodes for such nodes, and so can't easily set up dentries (we'd have to
have ghost inodes or something).
With this patch we don't actually create dentries until we get handles
from the server that we can use to set up their inodes, and we don't
actually bind them into the tree until we know for sure where they go.
(2) Inaccessible symbolic links.
If we're asked to mount two exports from the server, eg:
mount warthog:/warthog/aaa/xxx /mmm
mount warthog:/warthog/bbb/yyy /nnn
We may not be able to access anything nearer the root than xxx and yyy,
but we may find out later that /mmm/www/yyy, say, is actually the same
directory as the one mounted on /nnn. What we might then find out, for
example, is that /warthog/bbb was actually a symbolic link to
/warthog/aaa/xxx/www, but we can't actually determine that by talking to
the server until /warthog is made available by NFS.
This would lead to having constructed an errneous dentry tree which we
can't easily fix. We can end up with a dentry marked as a directory when
it should actually be a symlink, or we could end up with an apparently
hardlinked directory.
With this patch we need not make assumptions about the type of a dentry
for which we can't retrieve information, nor need we assume we know its
place in the grand scheme of things until we actually see that place.
This patch reduces the possibility of aliasing in the inode and page caches for
inodes that may be accessed by more than one NFS export. It also reduces the
number of superblocks required for NFS where there are many NFS exports being
used from a server (home directory server + autofs for example).
This in turn makes it simpler to do local caching of network filesystems, as it
can then be guaranteed that there won't be links from multiple inodes in
separate superblocks to the same cache file.
Obviously, cache aliasing between different levels of NFS protocol could still
be a problem, but at least that gives us another key to use when indexing the
cache.
This patch makes the following changes:
(1) The server record construction/destruction has been abstracted out into
its own set of functions to make things easier to get right. These have
been moved into fs/nfs/client.c.
All the code in fs/nfs/client.c has to do with the management of
connections to servers, and doesn't touch superblocks in any way; the
remaining code in fs/nfs/super.c has to do with VFS superblock management.
(2) The sequence of events undertaken by NFS mount is now reordered:
(a) A volume representation (struct nfs_server) is allocated.
(b) A server representation (struct nfs_client) is acquired. This may be
allocated or shared, and is keyed on server address, port and NFS
version.
(c) If allocated, the client representation is initialised. The state
member variable of nfs_client is used to prevent a race during
initialisation from two mounts.
(d) For NFS4 a simple pathwalk is performed, walking from FH to FH to find
the root filehandle for the mount (fs/nfs/getroot.c). For NFS2/3 we
are given the root FH in advance.
(e) The volume FSID is probed for on the root FH.
(f) The volume representation is initialised from the FSINFO record
retrieved on the root FH.
(g) sget() is called to acquire a superblock. This may be allocated or
shared, keyed on client pointer and FSID.
(h) If allocated, the superblock is initialised.
(i) If the superblock is shared, then the new nfs_server record is
discarded.
(j) The root dentry for this mount is looked up from the root FH.
(k) The root dentry for this mount is assigned to the vfsmount.
(3) nfs_readdir_lookup() creates dentries for each of the entries readdir()
returns; this function now attaches disconnected trees from alternate
roots that happen to be discovered attached to a directory being read (in
the same way nfs_lookup() is made to do for lookup ops).
The new d_materialise_unique() function is now used to do this, thus
permitting the whole thing to be done under one set of locks, and thus
avoiding any race between mount and lookup operations on the same
directory.
(4) The client management code uses a new debug facility: NFSDBG_CLIENT which
is set by echoing 1024 to /proc/net/sunrpc/nfs_debug.
(5) Clone mounts are now called xdev mounts.
(6) Use the dentry passed to the statfs() op as the handle for retrieving fs
statistics rather than the root dentry of the superblock (which is now a
dummy).
Signed-Off-By: David Howells <[email protected]>
Signed-off-by: Trond Myklebust <[email protected]>
Start rpciod in the server common (nfs_client struct) management code rather
than in the superblock management code. This means we only need to "start" it
once per server instead of once per superblock.
Signed-Off-By: David Howells <[email protected]>
Signed-off-by: Trond Myklebust <[email protected]>
Eliminate nfs_server::client_sys in favour of nfs_client::cl_rpcclient as we
only really need one per server that we're talking to since it doesn't have any
security on it.
The retransmission management variables are also moved to the common struct as
they're required to set up the cl_rpcclient connection.
The NFS2/3 client and client_acl connections are thenceforth derived by cloning
the cl_rpcclient connection and post-applying the authorisation flavour.
The code for setting up the initial common connection has been moved to
client.c as nfs_create_rpc_client(). All the NFS program definition tables are
also moved there as that's where they're now required rather than super.c.
Signed-Off-By: David Howells <[email protected]>
Signed-off-by: Trond Myklebust <[email protected]>
Move the rpc_ops from the nfs_server struct to the nfs_client struct as they're
common to all server records of a particular NFS protocol version.
Signed-Off-By: David Howells <[email protected]>
Signed-off-by: Trond Myklebust <[email protected]>
Make better use of inode* dereferencing macros to hide dereferencing chains
(including NFS_PROTO and NFS_CLIENT).
Signed-Off-By: David Howells <[email protected]>
Signed-off-by: Trond Myklebust <[email protected]>
Maintain a common server record for NFS2/3 as well as for NFS4 so that common
stuff can be moved there from struct nfs_server.
Signed-Off-By: David Howells <[email protected]>
Signed-off-by: Trond Myklebust <[email protected]>
Use the nominated dentry's superblock directly in the NFS statfs() op to get a
file handle, rather than using s_root (which will become a dummy dentry in a
future patch).
Signed-Off-By: David Howells <[email protected]>
Signed-off-by: Trond Myklebust <[email protected]>
Generalise the nfs_client structure by:
(1) Moving nfs_client to a more general place (nfs_fs_sb.h).
(2) Renaming its maintenance routines to be non-NFS4 specific.
(3) Move those maintenance routines to a new non-NFS4 specific file (client.c)
and move the declarations to internal.h.
(4) Make nfs_find/get_client() take a full sockaddr_in to include the port
number (will be required for NFS2/3).
(5) Make nfs_find/get_client() take the NFS protocol version (again will be
required to differentiate NFS2, 3 & 4 client records).
Also:
(6) Make nfs_client construction proceed akin to inodes, marking them as under
construction and providing a function to indicate completion.
(7) Make nfs_get_client() wait interruptibly if it finds a client that it can
share, but that client is currently being constructed.
(8) Make nfs4_create_client() use (6) and (7) instead of locking cl_sem.
Signed-Off-By: David Howells <[email protected]>
Signed-off-by: Trond Myklebust <[email protected]>
Add a set_capabilities NFS RPC op so that the server capabilities can be set.
Signed-Off-By: David Howells <[email protected]>
Signed-off-by: Trond Myklebust <[email protected]>
Add a lookup filehandle NFS RPC op so that a file handle can be looked up
without requiring dentries and inodes and other VFS stuff when doing an NFS4
pathwalk during mounting.
Signed-Off-By: David Howells <[email protected]>
Signed-off-by: Trond Myklebust <[email protected]>
Return an error when starting the idmapping pipe so that we can detect it
failing.
Signed-Off-By: David Howells <[email protected]>
Signed-off-by: Trond Myklebust <[email protected]>
Rename nfs_server::nfs4_state to nfs_client as it will be used to represent the
client state for NFS2 and NFS3 also.
Signed-Off-By: David Howells <[email protected]>
Signed-off-by: Trond Myklebust <[email protected]>
Rename struct nfs4_client to struct nfs_client so that it can become the basis
for a general client record for NFS2 and NFS3 in addition to NFS4.
Signed-Off-By: David Howells <[email protected]>
Signed-off-by: Trond Myklebust <[email protected]>
Make the nfs_callback_up()/down() prototypes just do nothing if NFS4 is not
enabled. Also make the down function void type since we can't really do
anything if it fails.
Signed-Off-By: David Howells <[email protected]>
Signed-off-by: Trond Myklebust <[email protected]>
Rename the NFS4 version of nfs_stat_to_errno() so that it doesn't conflict with
the common one used by NFS2 and NFS3.
Signed-Off-By: David Howells <[email protected]>
Signed-off-by: Trond Myklebust <[email protected]>
Fix ups for the splitting of the superblock stuff out of fs/nfs/inode.c,
including:
(*) Move the callback tcpport module param into callback.c.
(*) Move the idmap cache timeout module param into idmap.c.
(*) Changes to internal.h:
(*) namespace-nfs4.c was renamed to nfs4namespace.c.
(*) nfs_stat_to_errno() is in nfs2xdr.c, not nfs4xdr.c.
(*) nfs4xdr.c is contingent on CONFIG_NFS_V4.
(*) nfs4_path() is only uses if CONFIG_NFS_V4 is set.
Plus also:
(*) The sec_flavours[] table should really be const.
Signed-Off-By: David Howells <[email protected]>
Signed-off-by: Trond Myklebust <[email protected]>
The attached patch adds a new directory cache management function that prepares
a disconnected anonymous function to be connected into the dentry tree. The
anonymous dentry is transferred the name and parentage from another dentry.
The following changes were made in [try #2]:
(*) d_materialise_dentry() now switches the parentage of the two nodes around
correctly when one or other of them is self-referential.
The following changes were made in [try #7]:
(*) d_instantiate_unique() has had the interior part split out as function
__d_instantiate_unique(). Callers of this latter function must be holding
the appropriate locks.
(*) _d_rehash() has been added as a wrapper around __d_rehash() to call it
with the most obvious hash list (the one from the name). d_rehash() now
calls _d_rehash().
(*) d_materialise_dentry() is now __d_materialise_dentry() and is static.
(*) d_materialise_unique() added to perform the combination of d_find_alias(),
d_materialise_dentry() and d_add_unique() that the NFS client was doing
twice, all within a single dcache_lock critical section. This reduces the
number of times two different spinlocks were being accessed.
The following further changes were made:
(*) Add the dentries onto their parents d_subdirs lists.
Signed-Off-By: David Howells <[email protected]>
Signed-off-by: Trond Myklebust <[email protected]>
Export copy_4K_page() for use by modules via copy_page() (such as
CacheFiles).
Signed-Off-By: David Howells <[email protected]>
Signed-off-by: Paul Mackerras <[email protected]>
Check the bounds of length specifiers more thoroughly in the XDR decoding of
NFS4 readdir reply data.
Currently, if the server returns a bitmap or attr length that causes the
current decode point pointer to wrap, this could go undetected (consider a
small "negative" length on a 32-bit machine).
Also add a check into the main XDR decode handler to make sure that the amount
of data is a multiple of four bytes (as specified by RFC-1014). This makes
sure that we can do u32* pointer subtraction in the NFS client without risking
an undefined result (the result is undefined if the pointers are not correctly
aligned with respect to one another).
Signed-Off-By: David Howells <[email protected]>
Signed-off-by: Trond Myklebust <[email protected]>
(cherry picked from 5861fddd64a7eaf7e8b1a9997455a24e7f688092 commit)
Introduce the use of asm-offsets into the FRV architecture.
Signed-off-by: David Howells <[email protected]>
Signed-off-by: Andrew Morton <[email protected]>
Signed-off-by: Linus Torvalds <[email protected]>
Move the roundup() macro from binfmt_elf.c into linux/kernel.h as it's
generally useful.
[[email protected]: nuke all the other implementations]
Signed-off-by: David Howells <[email protected]>
Signed-off-by: Andrew Morton <[email protected]>
Signed-off-by: Linus Torvalds <[email protected]>
Add definitions for SEEK_SET, SEEK_CUR and SEEK_END to the kernel header
files.
Signed-off-by: David Howells <[email protected]>
Signed-off-by: Andrew Morton <[email protected]>
Signed-off-by: Linus Torvalds <[email protected]>
Adjust the ELF-FDPIC binfmt driver to conform much more to the CodingStyle,
silly though it may be.
Further changes:
(*) Drop the casts to long for addresses in kdebug() statements (they're
unsigned long already).
(*) Use extra variables to avoid expressions longer than 80 chars by splitting
the statement into multiple statements and letting the compiler optimise
them back together.
(*) Eliminate duplicate call of ksize() when working out how much space was
actually allocated for the stack.
(*) Discard the commented-out load_shlib prototype and op pointer as this will
not be supported in ELF-FDPIC for the foreseeable future.
Signed-off-by: David Howells <[email protected]>
Signed-off-by: Andrew Morton <[email protected]>
Signed-off-by: Linus Torvalds <[email protected]>
Fix execution through the FDPIC binfmt of programs stored on ramfs by
preventing the ramfs mmap() returning successfully on a private mapping of
a ramfs file. This causes NOMMU mmap to make a copy of the mapped portion
of the file and map that instead.
This could be improved by granting direct mapping access to read-only
private mappings for which the data is stored on a contiguous run of pages.
However, this is only likely to be the case if the file was extended with
truncate before being written.
ramfs is left to map the file directly for shared mappings so that SYSV IPC
and POSIX shared memory both still work.
Signed-off-by: David Howells <[email protected]>
Cc: Hugh Dickins <[email protected]>
Signed-off-by: Andrew Morton <[email protected]>
Signed-off-by: Linus Torvalds <[email protected]>
Fix some FRV arch compile errors, including:
(*) Marking nr_kernel_pages as __meminitdata so that references to it end up
being properly calculated rather than being assumed to be in the small
data section (and thus calculated wrt the GP register). Not doing this
causes the linker to emit errors as the offset is too big to fit into the
load instruction.
(*) Move pm_power_off into an unconditionally compiled .c file as it's now
unconditionally accessed.
(*) Declare frv_change_cmode() in a header file rather than in a .c file, and
declare it asmlinkage.
Signed-off-by: David Howells <[email protected]>
Signed-off-by: Andrew Morton <[email protected]>
Signed-off-by: Linus Torvalds <[email protected]>
The proposed NFS key type uses its own method of passing key requests to
userspace (upcalling) rather than invoking /sbin/request-key. This is
because the responsible userspace daemon should already be running and will
be contacted through rpc_pipefs.
This patch permits the NFS filesystem to pass auxiliary data to the upcall
operation (struct key_type::request_key) so that the upcaller can use a
pre-existing communications channel more easily.
Signed-off-by: David Howells <[email protected]>
Acked-By: Kevin Coffman <[email protected]>
Signed-off-by: Andrew Morton <[email protected]>
Signed-off-by: Linus Torvalds <[email protected]>
Cause the keys linked to a keyring to be unlinked from it when revoked and it
causes the data attached to a user-defined key to be discarded when revoked.
This frees up most of the quota a key occupied at that point, rather than
waiting for the key to actually be destroyed.
Signed-off-by: David Howells <[email protected]>
Signed-off-by: Andrew Morton <[email protected]>
Signed-off-by: Linus Torvalds <[email protected]>
Add the ability for key creation to overrun the user's quota in some
circumstances - notably when a session keyring is created and assigned to a
process that didn't previously have one.
This means it's still possible to log in, should PAM require the creation of a
new session keyring, and fix an overburdened key quota.
Signed-off-by: David Howells <[email protected]>
Signed-off-by: Andrew Morton <[email protected]>
Signed-off-by: Linus Torvalds <[email protected]>
Make another couple of alterations to the memory barrier document following
suggestions by Alan Stern and in co-operation with Paul McKenney:
(*) Rework the point of introduction of memory barriers and the description
of what they are to reiterate why they're needed.
(*) Modify a statement about the use of data dependency barriers to note that
other barriers can be used instead (as they imply DD-barriers).
Signed-off-by: David Howells <[email protected]>
Acked-By: Paul E. McKenney <[email protected]>
Signed-off-by: Andrew Morton <[email protected]>
Signed-off-by: Linus Torvalds <[email protected]>
Add a new VMA operation to notify a filesystem or other driver about the
MMU generating a fault because userspace attempted to write to a page
mapped through a read-only PTE.
This facility permits the filesystem or driver to:
(*) Implement storage allocation/reservation on attempted write, and so to
deal with problems such as ENOSPC more gracefully (perhaps by generating
SIGBUS).
(*) Delay making the page writable until the contents have been written to a
backing cache. This is useful for NFS/AFS when using FS-Cache/CacheFS.
It permits the filesystem to have some guarantee about the state of the
cache.
(*) Account and limit number of dirty pages. This is one piece of the puzzle
needed to make shared writable mapping work safely in FUSE.
Needed by cachefs (Or is it cachefiles? Or fscache? <head spins>).
At least four other groups have stated an interest in it or a desire to use
the functionality it provides: FUSE, OCFS2, NTFS and JFFS2. Also, things like
EXT3 really ought to use it to deal with the case of shared-writable mmap
encountering ENOSPC before we permit the page to be dirtied.
From: Peter Zijlstra <[email protected]>
get_user_pages(.write=1, .force=1) can generate COW hits on read-only
shared mappings, this patch traps those as mkpage_write candidates and fails
to handle them the old way.
Signed-off-by: David Howells <[email protected]>
Cc: Miklos Szeredi <[email protected]>
Cc: Joel Becker <[email protected]>
Cc: Mark Fasheh <[email protected]>
Cc: Anton Altaparmakov <[email protected]>
Cc: David Woodhouse <[email protected]>
Cc: Hugh Dickins <[email protected]>
Signed-off-by: Andrew Morton <[email protected]>
Signed-off-by: Linus Torvalds <[email protected]>
Enable XFS to limit the statfs() results to the project quota covering the
dentry used as a base for call.
Signed-off-by: David Howells <[email protected]>
Signed-off-by: Nathan Scott <[email protected]>
Signed-off-by: Andrew Morton <[email protected]>
Signed-off-by: Linus Torvalds <[email protected]>
Give the statfs superblock operation a dentry pointer rather than a superblock
pointer.
This complements the get_sb() patch. That reduced the significance of
sb->s_root, allowing NFS to place a fake root there. However, NFS does
require a dentry to use as a target for the statfs operation. This permits
the root in the vfsmount to be used instead.
linux/mount.h has been added where necessary to make allyesconfig build
successfully.
Interest has also been expressed for use with the FUSE and XFS filesystems.
Signed-off-by: David Howells <[email protected]>
Acked-by: Al Viro <[email protected]>
Cc: Nathan Scott <[email protected]>
Signed-off-by: Andrew Morton <[email protected]>
Signed-off-by: Linus Torvalds <[email protected]>
Extend the get_sb() filesystem operation to take an extra argument that
permits the VFS to pass in the target vfsmount that defines the mountpoint.
The filesystem is then required to manually set the superblock and root dentry
pointers. For most filesystems, this should be done with simple_set_mnt()
which will set the superblock pointer and then set the root dentry to the
superblock's s_root (as per the old default behaviour).
The get_sb() op now returns an integer as there's now no need to return the
superblock pointer.
This patch permits a superblock to be implicitly shared amongst several mount
points, such as can be done with NFS to avoid potential inode aliasing. In
such a case, simple_set_mnt() would not be called, and instead the mnt_root
and mnt_sb would be set directly.
The patch also makes the following changes:
(*) the get_sb_*() convenience functions in the core kernel now take a vfsmount
pointer argument and return an integer, so most filesystems have to change
very little.
(*) If one of the convenience function is not used, then get_sb() should
normally call simple_set_mnt() to instantiate the vfsmount. This will
always return 0, and so can be tail-called from get_sb().
(*) generic_shutdown_super() now calls shrink_dcache_sb() to clean up the
dcache upon superblock destruction rather than shrink_dcache_anon().
This is required because the superblock may now have multiple trees that
aren't actually bound to s_root, but that still need to be cleaned up. The
currently called functions assume that the whole tree is rooted at s_root,
and that anonymous dentries are not the roots of trees which results in
dentries being left unculled.
However, with the way NFS superblock sharing are currently set to be
implemented, these assumptions are violated: the root of the filesystem is
simply a dummy dentry and inode (the real inode for '/' may well be
inaccessible), and all the vfsmounts are rooted on anonymous[*] dentries
with child trees.
[*] Anonymous until discovered from another tree.
(*) The documentation has been adjusted, including the additional bit of
changing ext2_* into foo_* in the documentation.
[[email protected]: convert ipath_fs, do other stuff]
Signed-off-by: David Howells <[email protected]>
Acked-by: Al Viro <[email protected]>
Cc: Nathan Scott <[email protected]>
Cc: Roland Dreier <[email protected]>
Signed-off-by: Andrew Morton <[email protected]>
Signed-off-by: Linus Torvalds <[email protected]>
Add a revocation notification method to the key type and calls it whilst
the key's semaphore is still write-locked after setting the revocation
flag.
The patch then uses this to maintain a reference on the task_struct of the
process that calls request_key() for as long as the authorisation key
remains unrevoked.
This fixes a potential race between two processes both of which have
assumed the authority to instantiate a key (one may have forked the other
for example). The problem is that there's no locking around the check for
revocation of the auth key and the use of the task_struct it points to, nor
does the auth key keep a reference on the task_struct.
Access to the "context" pointer in the auth key must thenceforth be done
with the auth key semaphore held. The revocation method is called with the
target key semaphore held write-locked and the search of the context
process's keyrings is done with the auth key semaphore read-locked.
The check for the revocation state of the auth key just prior to searching
it is done after the auth key is read-locked for the search. This ensures
that the auth key can't be revoked between the check and the search.
The revocation notification method is added so that the context task_struct
can be released as soon as instantiation happens rather than waiting for
the auth key to be destroyed, thus avoiding the unnecessary pinning of the
requesting process.
Signed-off-by: David Howells <[email protected]>
Signed-off-by: Andrew Morton <[email protected]>
Signed-off-by: Linus Torvalds <[email protected]>
From: David Howells <[email protected]>
Apply some alterations to the memory barrier document that I worked out
with Paul McKenney of IBM, plus some of the alterations suggested by Alan
Stern.
The following changes were made:
(*) One of the examples given for what can happen with overlapping memory
barriers was wrong.
(*) The description of general memory barriers said that a general barrier is
a combination of a read barrier and a write barrier. This isn't entirely
true: it implies both, but is more than a combination of both.
(*) The first example in the "SMP Barrier Pairing" section was wrong: the
loads around the read barrier need to touch the memory locations in the
opposite order to the stores around the write barrier.
(*) Added a note to make explicit that the loads should be in reverse order to
the stores.
(*) Adjusted the diagrams in the "Examples Of Memory Barrier Sequences"
section to make them clearer. Added a couple of diagrams to make it more
clear as to how it could go wrong without the barrier.
(*) Added a section on memory speculation.
(*) Dropped any references to memory allocation routines doing memory
barriers. They may do sometimes, but it can't be relied on. This may be
worthy of further documentation later.
(*) Made the fact that a LOCK followed by an UNLOCK should not be considered a
full memory barrier more explicit and gave an example.
Signed-off-by: David Howells <[email protected]>
Acked-by: Paul E. McKenney <[email protected]>
Signed-off-by: Andrew Morton <[email protected]>
Signed-off-by: Linus Torvalds <[email protected]>
As fs/nfs/inode.c is rather large, heterogenous and unwieldy, the attached
patch splits it up into a number of files:
(*) fs/nfs/inode.c
Strictly inode specific functions.
(*) fs/nfs/super.c
Superblock management functions for NFS and NFS4, normal access, clones
and referrals. The NFS4 superblock functions _could_ move out into a
separate conditionally compiled file, but it's probably not worth it as
there're so many common bits.
(*) fs/nfs/namespace.c
Some namespace-specific functions have been moved here.
(*) fs/nfs/nfs4namespace.c
NFS4-specific namespace functions (this could be merged into the previous
file). This file is conditionally compiled.
(*) fs/nfs/internal.h
Inter-file declarations, plus a few simple utility functions moved from
fs/nfs/inode.c.
Additionally, all the in-.c-file externs have been moved here, and those
files they were moved from now includes this file.
For the most part, the functions have not been changed, only some multiplexor
functions have changed significantly.
I've also:
(*) Added some extra banner comments above some functions.
(*) Rearranged the function order within the files to be more logical and
better grouped (IMO), though someone may prefer a different order.
(*) Reduced the number of #ifdefs in .c files.
(*) Added missing __init and __exit directives.
Signed-Off-By: David Howells <[email protected]>
Remove an unnecessary memory barrier (implicit in rcu_dereference()) from
install_session_keyring().
install_session_keyring() is also rearranged a little to make it slightly
more efficient.
As install_*_keyring() may schedule (in synchronize_rcu() or
keyring_alloc()), they may not be entered with interrupts disabled - and so
there's no point saving the interrupt disablement state over the critical
section.
exec_keys() will also be invoked with interrupts enabled, and so that doesn't
need to save the interrupt state either.
Signed-off-by: David Howells <[email protected]>
Signed-off-by: Andrew Morton <[email protected]>
Signed-off-by: Linus Torvalds <[email protected]>
In the memory barrier document, improve the example of the data dependency
barrier situation by:
(1) showing the initial values of the variables involved; and
(2) repeating the instruction sequence description, this time with the data
dependency barrier actually shown to make it clear what the revised
sequence actually is.
Signed-off-by: David Howells <[email protected]>
Signed-off-by: Andrew Morton <[email protected]>
Signed-off-by: Linus Torvalds <[email protected]>
Fix the memory barrier documentation to attempt to describe atomic ops
correctly.
atomic_t ops that return a value _do_ imply smp_mb() either side, and so
don't actually require smp_mb__*_atomic_*() special barriers.
Also explains why special barriers exist in addition to normal barriers.
Further fix the memory barrier documents to portray bitwise operation
memory barrier effects correctly following Nick Piggin's comments.
It makes the point that any atomic op that both modifies some state in
memory and returns information on that state implies memory barriers on
both sides.
Signed-off-by: David Howells <[email protected]>
Signed-off-by: Andrew Morton <[email protected]>
Signed-off-by: Linus Torvalds <[email protected]>
This fixes the problem of an oops occuring when a user attempts to add a
key to a non-keyring key [CVE-2006-1522].
The problem is that __keyring_search_one() doesn't check that the
keyring it's been given is actually a keyring.
I've fixed this problem by:
(1) declaring that caller of __keyring_search_one() must guarantee that
the keyring is a keyring; and
(2) making key_create_or_update() check that the keyring is a keyring,
and return -ENOTDIR if it isn't.
This can be tested by:
keyctl add user b b `keyctl add user a a @s`
Signed-off-by: David Howells <[email protected]>
Signed-off-by: Linus Torvalds <[email protected]>
The attached patch documents the Linux kernel's memory barriers.
I've updated it from the comments I've been given.
The per-arch notes sections are gone because it's clear that there are so many
exceptions, that it's not worth having them.
I've added a list of references to other documents.
I've tried to get rid of the concept of memory accesses appearing on the bus;
what matters is apparent behaviour with respect to other observers in the
system.
Interrupts barrier effects are now considered to be non-existent. They may be
there, but you may not rely on them.
I've added a couple of definition sections at the top of the document: one to
specify the minimum execution model that may be assumed, the other to specify
what this document refers to by the term "memory".
I've made greater mention of the use of mmiowb().
I've adjusted the way in which caches are described, and described the fun
that can be had with cache coherence maintenance being unordered and data
dependency not being necessarily implicit.
I've described (smp_)read_barrier_depends().
I've rearranged the order of the sections, so that memory barriers are
discussed in abstract first, and then described the memory barrier facilities
available on Linux, before going on to more real-world discussions and examples.
I've added information about the lack of memory barriering effects with atomic
ops and bitops.
I've added information about control dependencies.
I've added more diagrams to illustrate caching interactions between CPUs.
Signed-off-by: David Howells <[email protected]>
Signed-off-by: Linus Torvalds <[email protected]>
Cause an attempt to add a duplicate non-updateable key (such as a keyring) to
a keyring to discard the extant copy in favour of the new one rather than
failing with EEXIST:
# do the test in an empty session
keyctl session
# create a new keyring called "a" and attach to session
keyctl newring a @s
# create another new keyring called "a" and attach to session,
# displacing the keyring added by the second command:
keyctl newring a @s
Without this patch, the third command will fail.
For updateable keys (such as those of "user" type), the update method will
still be called rather than a new key being created.
Signed-off-by: David Howells <[email protected]>
Signed-off-by: Andrew Morton <[email protected]>
Signed-off-by: Linus Torvalds <[email protected]>
David Howells