- use kzalloc() instead of kmalloc() + memset()
- improve a printk info
Signed-off-by: Li Zefan <lizf@cn.fujitsu.com>
Signed-off-by: Theodore Ts'o <tytso@mit.edu>
The truncate patch should not use the i_allocated_meta_blocks
value. So add seperate functions to be used in the truncate
and alloc path. We also need to release the meta-data block
that we reserved for the blocks that we are truncating.
Signed-off-by: Aneesh Kumar K.V <aneesh.kumar@linux.vnet.ibm.com>
Signed-off-by: Theodore Ts'o <tytso@mit.edu>
With the FLEX_BG layout, there is no reason why extents can't cross
block groups, so make the truncate code reserve enough credits so we
don't BUG if we come across such an extent.
Signed-off-by: "Theodore Ts'o" <tytso@mit.edu>
The extents codepath for ext4_truncate() requests journal transaction
credits in very small chunks, requesting only what is needed. This
means there may not be enough credits left on the transaction handle
after ext4_truncate() returns and then when ext4_delete_inode() tries
finish up its work, it may not have enough transaction credits,
causing a BUG() oops in the jbd2 core.
Also, reserve an extra 2 blocks when starting an ext4_delete_inode()
since we need to update the inode bitmap, as well as update the
orphaned inode linked list.
Signed-off-by: "Theodore Ts'o" <tytso@mit.edu>
The ext4_ext_journal_restart() is a convenience function which checks
to see if the requested number of credits is present, and if so it
closes the current transaction and attaches the current handle to the
new transaction. Unfortunately, it wasn't proprely checking the
return value from ext4_journal_extend(), so it was starting a new
transaction when one was not necessary, and returning an error when
all that was necessary was to restart the handle with a new
transaction.
Signed-off-by: "Theodore Ts'o" <tytso@mit.edu>
ext4_da_write_begin needs to call journal_stop before returning,
if the page allocation fails.
Signed-off-by: Eric Sandeen <sandeen@redhat.com>
Acked-by: Mingming Cao <cmm@us.ibm.com>
Signed-off-by: "Theodore Ts'o" <tytso@mit.edu>
In ordered mode, the current jbd2 aborts the journal if a file data buffer
has an error. But this behavior is unintended, and we found that it has
been adopted accidentally.
This patch undoes it and just calls printk() instead of aborting the
journal. Unlike a similar patch for ext3/jbd, file data buffers are
written via generic_writepages(). But we also need to set AS_EIO
into their mappings because wait_on_page_writeback_range() clears
AS_EIO before a user process sees it.
Signed-off-by: Hidehiro Kawai <hidehiro.kawai.ez@hitachi.com>
Signed-off-by: "Theodore Ts'o" <tytso@mit.edu>
A transient I/O error can corrupt inode data. Here is the scenario:
(1) update inode_A at the block_B
(2) pdflush writes out new inode_A to the filesystem, but it results
in write I/O error, at this point, BH_Uptodate flag of the buffer
for block_B is cleared and BH_Write_EIO is set
(3) create new inode_C which located at block_B, and
__ext4_get_inode_loc() tries to read on-disk block_B because the
buffer is not uptodate
(4) if it can read on-disk block_B successfully, inode_A is
overwritten by old data
This patch makes __ext4_get_inode_loc() not read the inode block if the
buffer has BH_Write_EIO flag. In this case, the buffer should have the
latest information, so setting the uptodate flag to the buffer (this
avoids WARN_ON_ONCE() in mark_buffer_dirty().)
According to this change, we would need to test BH_Write_EIO flag for the
error checking. Currently nobody checks write I/O errors on metadata
buffers, but it will be done in other patches I'm working on.
Signed-off-by: Hidehiro Kawai <hidehiro.kawai.ez@hitachi.com>
Cc: sugita <yumiko.sugita.yf@hitachi.com>
Cc: Satoshi OSHIMA <satoshi.oshima.fk@hitachi.com>
Cc: Nick Piggin <nickpiggin@yahoo.com.au>
Cc: Jan Kara <jack@ucw.cz>
Cc: <linux-ext4@vger.kernel.org>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Theodore Ts'o <tytso@mit.edu>
Currently, the locality group prealloc list is freed only when there
is a block allocation failure. This can result in large number of
entries in the preallocation list making ext4_mb_use_preallocated()
expensive.
To fix this, we convert the locality group prealloc list to a hash
list. The hash index is the order of number of blocks in the prealloc
space with a max order of 9. When adding prealloc space to the list we
make sure total entries for each order does not exceed 8. If it is
more than 8 we discard few entries and make sure the we have only <= 5
entries.
Signed-off-by: Aneesh Kumar K.V <aneesh.kumar@linux.vnet.ibm.com>
Signed-off-by: Theodore Ts'o <tytso@mit.edu>
NR_CPUS can be really large. We should be using nr_cpu_ids instead.
Signed-off-by: Aneesh Kumar K.V <aneesh.kumar@linux.vnet.ibm.com>
Signed-off-by: Theodore Ts'o <tytso@mit.edu>
Don't call BUG_ON on file system failures. Instead use ext4_error and
also handle the continue case properly.
Signed-off-by: Aneesh Kumar K.V <aneesh.kumar@linux.vnet.ibm.com>
Signed-off-by: Theodore Ts'o <tytso@mit.edu>
I noticed when filling a 1T filesystem with 4 threads using the
fs_mark benchmark:
fs_mark -d /mnt/test -D 256 -n 100000 -t 4 -s 20480 -F -S 0
that I occasionally got checksum mismatch errors:
EXT4-fs error (device sdb): ext4_init_inode_bitmap: Checksum bad for group 6935
etc. I'd reliably get 4-5 of them during the run.
It appears that the problem is likely a race to init the bg's
when the uninit_bg feature is enabled.
With the patch below, which adds sb_bgl_locking around initialization,
I was able to complete several runs with no errors or warnings.
Signed-off-by: Eric Sandeen <sandeen@redhat.com>
Signed-off-by: Theodore Ts'o <tytso@mit.edu>
ext4_read_block_bitmap and read_inode_bitmap do essentially
the same thing, and yet they are structured quite differently.
I came across this difference while looking at doing bg locking
during bg initialization.
This patch:
* removes unnecessary casts in the error messages
* renames read_inode_bitmap to ext4_read_inode_bitmap
* and more substantially, restructures the inode bitmap
reading function to be more like the block bitmap counterpart.
The change to the inode bitmap reader simplifies the locking
to be applied in the next patch.
Signed-off-by: Eric Sandeen <sandeen@redhat.com>
Signed-off-by: Theodore Ts'o <tytso@mit.edu>
If the block group checksums are corrupted, still allow the mount to
succeed, so e2fsck can have a chance to try to fix things up. Add
code in the remount r/w path to make sure the block group checksums
are valid before allowing the filesystem to be remounted read/write.
Signed-off-by: "Theodore Ts'o" <tytso@mit.edu>
Inserting an extent can cause a new entry in the already existing index
block. That doesn't increase the depth of the instead. Instead it adds a
new leaf block. Now with the new leaf block the path information
corresponding to the logical block should be fetched from the new block.
The old path will be pointing to the old leaf block.
We need to recalucate the path information on extent insert
even if depth doesn't change. Without this change, the extent merge
after converting an unwritten extent to initialized extent takes the wrong
extent and cause data corruption.
Signed-off-by: Aneesh Kumar K.V <aneesh.kumar@linux.vnet.ibm.com>
Signed-off-by: Mingming Cao <cmm@us.ibm.com>
Signed-off-by: "Theodore Ts'o" <tytso@mit.edu>
* git://git.kernel.org/pub/scm/linux/kernel/git/rusty/linux-2.6-for-linus:
lguest: turn Waker into a thread, not a process
lguest: Enlarge virtio rings
lguest: Use GSO/IFF_VNET_HDR extensions on tun/tap
lguest: Remove 'network: no dma buffer!' warning
lguest: Adaptive timeout
lguest: Tell Guest net not to notify us on every packet xmit
lguest: net block unneeded receive queue update notifications
lguest: wrap last_avail accesses.
lguest: use cpu capability accessors
lguest: virtio-rng support
lguest: Support assigning a MAC address
lguest: Don't leak /dev/zero fd
lguest: fix verbose printing of device features.
lguest: fix switcher_page leak on unload
lguest: Guest int3 fix
lguest: set max_pfn_mapped, growl loudly at Yinghai Lu
* 'for-linus' of git://git.o-hand.com/linux-mfd:
mfd: accept pure device as a parent, not only platform_device
mfd: add platform_data to mfd_cell
mfd: Coding style fixes
mfd: Use to_platform_device instead of container_of
* 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/jbarnes/pci-2.6: (21 commits)
x86/PCI: use dev_printk when possible
PCI: add D3 power state avoidance quirk
PCI: fix bogus "'device' may be used uninitialized" warning in pci_slot
PCI: add an option to allow ASPM enabled forcibly
PCI: disable ASPM on pre-1.1 PCIe devices
PCI: disable ASPM per ACPI FADT setting
PCI MSI: Don't disable MSIs if the mask bit isn't supported
PCI: handle 64-bit resources better on 32-bit machines
PCI: rewrite PCI BAR reading code
PCI: document pci_target_state
PCI hotplug: fix typo in pcie hotplug output
x86 gart: replace to_pages macro with iommu_num_pages
x86, AMD IOMMU: replace to_pages macro with iommu_num_pages
iommu: add iommu_num_pages helper function
dma-coherent: add documentation to new interfaces
Cris: convert to using generic dma-coherent mem allocator
Sh: use generic per-device coherent dma allocator
ARM: support generic per-device coherent dma mem
Generic dma-coherent: fix DMA_MEMORY_EXCLUSIVE
x86: use generic per-device dma coherent allocator
...
Alexey Dobriyan reported trouble with LTP with the new fast-gup code,
and Johannes Weiner debugged it to non-page-aligned addresses, where the
new get_user_pages_fast() code would do all the wrong things, including
just traversing past the end of the requested area due to 'addr' never
matching 'end' exactly.
This is not a pretty fix, and we may actually want to move the alignment
into generic code, leaving just the core code per-arch, but Alexey
verified that the vmsplice01 LTP test doesn't crash with this.
Reported-and-tested-by: Alexey Dobriyan <adobriyan@gmail.com>
Debugged-by: Johannes Weiner <hannes@saeurebad.de>
Cc: Nick Piggin <npiggin@suse.de>
Cc: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
lguest uses a Waker process to break it out of the kernel (ie.
actually running the guest) when file descriptor needs attention.
Changing this from a process to a thread somewhat simplifies things:
it can directly access the fd_set of things to watch. More
importantly, it means that the Waker can see Guest memory correctly,
so /dev/vring file descriptors will work as anticipated (the
alternative is to actually mmap MAP_SHARED, but you can't do that with
/dev/zero).
Signed-off-by: Rusty Russell <rusty@rustcorp.com.au>
Since the correct timeout value varies, use a heuristic which adjusts
the timeout depending on how many packets we've seen. This gives
slightly worse results, but doesn't need tweaking when GSO is
introduced.
500 usec 19.1887 xmit 561141 recv 1 timeout 559657
Dynamic (278) 20.1974 xmit 214510 recv 5 timeout 214491 usec 278
Signed-off-by: Rusty Russell <rusty@rustcorp.com.au>
virtio_ring has the ability to suppress notifications. This prevents
a guest exit for every packet, but we need to set a timer on packet
receipt to re-check if there were any remaining packets.
Here are the times for 1G TCP Guest->Host with different timeout
settings (it matters because the TCP window doesn't grow big enough to
fill the entire buffer):
Timeout value Seconds Xmit/Recv/Timeout
None (before) 25.3784 xmit 7750233 recv 1
2500 usec 62.5119 xmit 207020 recv 2 timeout 207020
1000 usec 34.5379 xmit 207003 recv 2 timeout 207003
750 usec 29.2305 xmit 207002 recv 1 timeout 207002
500 usec 19.1887 xmit 561141 recv 1 timeout 559657
250 usec 20.0465 xmit 214128 recv 2 timeout 214110
100 usec 19.2583 xmit 561621 recv 1 timeout 560153
(Note that these values are sensitive to the GSO patches which come
later, and probably other traffic-related variables, so take with a
large grain of salt).
Signed-off-by: Rusty Russell <rusty@rustcorp.com.au>
To simplify the transition to when we publish indices in the ring
(and make shuffling my patch queue easier), wrap them in a lg_last_avail()
macro.
Signed-off-by: Rusty Russell <rusty@rustcorp.com.au>
To support my little make-x86-bitops-use-proper-typechecking projectlet.
Cc: Thomas Gleixner <tglx@linutronix.de>
Cc: Andrea Arcangeli <andrea@qumranet.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Acked-by: Ingo Molnar <mingo@elte.hu>
Signed-off-by: Rusty Russell <rusty@rustcorp.com.au>
This is a simple patch to add support for the virtio "hardware random
generator" to lguest. It gets about 1.2 MB/sec reading from /dev/hwrng
in the guest.
Signed-off-by: Rusty Russell <rusty@rustcorp.com.au>
If you've got a nice DHCP configuration which maps MAC
addresses to specific IP addresses, then you're going to
want to start your guest with one of those MAC addresses.
Also, in Fedora, we have persistent network interface naming
based on the MAC address, so with randomly assigned
addresses you're soon going to hit eth13. Who knows what
will happen then!
Allow assigning a MAC address to the network interface with
e.g.
--tunnet=bridge:eth0:00:FF:95:6B:DA:3D
or:
--tunnet=192.168.121.1:00:FF:95:6B:DA:3D
which is pretty unintelligable, but ...
(includes Rusty's minor rework)
Signed-off-by: Mark McLoughlin <markmc@redhat.com>
Signed-off-by: Rusty Russell <rusty@rustcorp.com.au>
map_switcher allocates the array, unmap_switcher has to free it
accordingly.
Signed-off-by: Johannes Weiner <hannes@saeurebad.de>
Signed-off-by: Rusty Russell <rusty@rustcorp.com.au>
Ron Minnich noticed that guest userspace gets a GPF when it tries to int3:
we need to copy the privilege level from the guest-supplied IDT to the real
IDT. int3 is the only common case where guest userspace expects to invoke
an interrupt, so that's the symptom of failing to do this.
Signed-off-by: Rusty Russell <rusty@rustcorp.com.au>
6af61a7614 'x86: clean up max_pfn_mapped
usage - 32-bit' makes the following comment:
XEN PV and lguest may need to assign max_pfn_mapped too.
But no CC. Yinghai, wasting fellow developers' time is a VERY bad
habit. If you do it again, I will hunt you down and try to extract
the three hours of my life I just lost :)
Signed-off-by: Rusty Russell <rusty@rustcorp.com.au>
Cc: Yinghai Lu <yhlu.kernel@gmail.com>
The computed color value is never actually written to hardware
colormap register.
Signed-off-by: Manuel Lauss <mano@roarinelk.homelinux.net>
Cc: Nobuhiro Iwamatsu <iwamatsu.nobuhiro@renesas.com>
Cc: Munakata Hisao <munakata.hisao@renesas.com>
Cc: Paul Mundt <lethal@linux-sh.org>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
With SLUB debugging turned on in 2.6.26, I was getting memory corruption
when testing eCryptfs. The root cause turned out to be that eCryptfs was
doing kmalloc(PAGE_CACHE_SIZE); virt_to_page() and treating that as a nice
page-aligned chunk of memory. But at least with SLUB debugging on, this
is not always true, and the page we get from virt_to_page does not
necessarily match the PAGE_CACHE_SIZE worth of memory we got from kmalloc.
My simple testcase was 2 loops doing "rm -f fileX; cp /tmp/fileX ." for 2
different multi-megabyte files. With this change I no longer see the
corruption.
Signed-off-by: Eric Sandeen <sandeen@redhat.com>
Acked-by: Michael Halcrow <mhalcrow@us.ibm.com>
Acked-by: Rik van Riel <riel@redhat.com>
Cc: <stable@kernel.org> [2.6.25.x, 2.6.26.x]
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
I got section mismatch message about bio_integrity_init_slab().
WARNING: fs/built-in.o(__ksymtab+0xb60): Section mismatch in reference from the variable __ksymtab_bio_integrity_init_slab to the function .init.text:bio_integrity_init_slab()
The symbol bio_integrity_init_slab is exported and annotated __init Fix
this by removing the __init annotation of bio_integrity_init_slab or drop
the export.
It only call from init_bio(). The EXPORT_SYMBOL() can be removed.
Signed-off-by: Yoichi Yuasa <yoichi_yuasa@tripeaks.co.jp>
Cc: "Martin K. Petersen" <martin.petersen@oracle.com>
Cc: Jens Axboe <jens.axboe@oracle.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
When we read some part of a file through pagecache, if there is a
pagecache of corresponding index but this page is not uptodate, read IO
is issued and this page will be uptodate.
I think this is good for pagesize == blocksize environment but there is
room for improvement on pagesize != blocksize environment. Because in
this case a page can have multiple buffers and even if a page is not
uptodate, some buffers can be uptodate.
So I suggest that when all buffers which correspond to a part of a file
that we want to read are uptodate, use this pagecache and copy data from
this pagecache to user buffer even if a page is not uptodate. This can
reduce read IO and improve system throughput.
I wrote a benchmark program and got result number with this program.
This benchmark do:
1: mount and open a test file.
2: create a 512MB file.
3: close a file and umount.
4: mount and again open a test file.
5: pwrite randomly 300000 times on a test file. offset is aligned
by IO size(1024bytes).
6: measure time of preading randomly 100000 times on a test file.
The result was:
2.6.26
330 sec
2.6.26-patched
226 sec
Arch:i386
Filesystem:ext3
Blocksize:1024 bytes
Memory: 1GB
On ext3/4, a file is written through buffer/block. So random read/write
mixed workloads or random read after random write workloads are optimized
with this patch under pagesize != blocksize environment. This test result
showed this.
The benchmark program is as follows:
#include <stdio.h>
#include <sys/types.h>
#include <sys/stat.h>
#include <fcntl.h>
#include <unistd.h>
#include <time.h>
#include <stdlib.h>
#include <string.h>
#include <sys/mount.h>
#define LEN 1024
#define LOOP 1024*512 /* 512MB */
main(void)
{
unsigned long i, offset, filesize;
int fd;
char buf[LEN];
time_t t1, t2;
if (mount("/dev/sda1", "/root/test1/", "ext3", 0, 0) < 0) {
perror("cannot mount\n");
exit(1);
}
memset(buf, 0, LEN);
fd = open("/root/test1/testfile", O_CREAT|O_RDWR|O_TRUNC);
if (fd < 0) {
perror("cannot open file\n");
exit(1);
}
for (i = 0; i < LOOP; i++)
write(fd, buf, LEN);
close(fd);
if (umount("/root/test1/") < 0) {
perror("cannot umount\n");
exit(1);
}
if (mount("/dev/sda1", "/root/test1/", "ext3", 0, 0) < 0) {
perror("cannot mount\n");
exit(1);
}
fd = open("/root/test1/testfile", O_RDWR);
if (fd < 0) {
perror("cannot open file\n");
exit(1);
}
filesize = LEN * LOOP;
for (i = 0; i < 300000; i++){
offset = (random() % filesize) & (~(LEN - 1));
pwrite(fd, buf, LEN, offset);
}
printf("start test\n");
time(&t1);
for (i = 0; i < 100000; i++){
offset = (random() % filesize) & (~(LEN - 1));
pread(fd, buf, LEN, offset);
}
time(&t2);
printf("%ld sec\n", t2-t1);
close(fd);
if (umount("/root/test1/") < 0) {
perror("cannot umount\n");
exit(1);
}
}
Signed-off-by: Hisashi Hifumi <hifumi.hisashi@oss.ntt.co.jp>
Cc: Nick Piggin <nickpiggin@yahoo.com.au>
Cc: Christoph Hellwig <hch@infradead.org>
Cc: Jan Kara <jack@ucw.cz>
Cc: <linux-ext4@vger.kernel.org>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
The original "Pass the bus number we expect the S3C24XX SPI driver to
attach to via the platform data." [1] patch was mis-sent, and missed two
important parts of the diff, which was to actually set the bus_num field
and add the relevant field to the platform data.
The previous commit 50f426b55d promised to
add a bus_num field, but failed to include the two hunks that added this
field to include/asm-arm/arch-s3c2410/spi.h and then pass it to the spi
core when creating the new master field in drivers/spi/spi_s3c24xx.c.
[1] git commit 50f426b55d
Signed-off-by: Ben Dooks <ben-linux@fluff.org>
Signed-off-by: David Brownell <dbrownell@users.sourceforge.net>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
The block transfer routine in the mpc52xx psc spi driver misinterpret
the datasheet. According to the processor datasheet the chipselect is
held as long as the EOF is not written.
Theoretically blocks of any sizes can be transferred in this way. The
old routine however writes an EOF after every word, which has the size
of size_of_word. This makes the transfer slow.
Also fixed some duplicate code.
Signed-off-by: Luotao Fu <l.fu@pengutronix.de>
Signed-off-by: David Brownell <dbrownell@users.sourceforge.net>
Cc: <stable@kernel.org> [2.6.25.x, 2.6.26.x]
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
This patch fixes the following build error on sh caused by commit
aa888a7497 ("hugetlb: support larger than
MAX_ORDER"):
mm/hugetlb.c: In function 'alloc_bootmem_huge_page':
mm/hugetlb.c:958: error: implicit declaration of function 'virt_to_phys'
Signed-off-by: Adrian Bunk <bunk@kernel.org>
Cc: Hirokazu Takata <takata@linux-m32r.org>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
With KVM/GFP/XPMEM there isn't just the primary CPU MMU pointing to pages.
There are secondary MMUs (with secondary sptes and secondary tlbs) too.
sptes in the kvm case are shadow pagetables, but when I say spte in
mmu-notifier context, I mean "secondary pte". In GRU case there's no
actual secondary pte and there's only a secondary tlb because the GRU
secondary MMU has no knowledge about sptes and every secondary tlb miss
event in the MMU always generates a page fault that has to be resolved by
the CPU (this is not the case of KVM where the a secondary tlb miss will
walk sptes in hardware and it will refill the secondary tlb transparently
to software if the corresponding spte is present). The same way
zap_page_range has to invalidate the pte before freeing the page, the spte
(and secondary tlb) must also be invalidated before any page is freed and
reused.
Currently we take a page_count pin on every page mapped by sptes, but that
means the pages can't be swapped whenever they're mapped by any spte
because they're part of the guest working set. Furthermore a spte unmap
event can immediately lead to a page to be freed when the pin is released
(so requiring the same complex and relatively slow tlb_gather smp safe
logic we have in zap_page_range and that can be avoided completely if the
spte unmap event doesn't require an unpin of the page previously mapped in
the secondary MMU).
The mmu notifiers allow kvm/GRU/XPMEM to attach to the tsk->mm and know
when the VM is swapping or freeing or doing anything on the primary MMU so
that the secondary MMU code can drop sptes before the pages are freed,
avoiding all page pinning and allowing 100% reliable swapping of guest
physical address space. Furthermore it avoids the code that teardown the
mappings of the secondary MMU, to implement a logic like tlb_gather in
zap_page_range that would require many IPI to flush other cpu tlbs, for
each fixed number of spte unmapped.
To make an example: if what happens on the primary MMU is a protection
downgrade (from writeable to wrprotect) the secondary MMU mappings will be
invalidated, and the next secondary-mmu-page-fault will call
get_user_pages and trigger a do_wp_page through get_user_pages if it
called get_user_pages with write=1, and it'll re-establishing an updated
spte or secondary-tlb-mapping on the copied page. Or it will setup a
readonly spte or readonly tlb mapping if it's a guest-read, if it calls
get_user_pages with write=0. This is just an example.
This allows to map any page pointed by any pte (and in turn visible in the
primary CPU MMU), into a secondary MMU (be it a pure tlb like GRU, or an
full MMU with both sptes and secondary-tlb like the shadow-pagetable layer
with kvm), or a remote DMA in software like XPMEM (hence needing of
schedule in XPMEM code to send the invalidate to the remote node, while no
need to schedule in kvm/gru as it's an immediate event like invalidating
primary-mmu pte).
At least for KVM without this patch it's impossible to swap guests
reliably. And having this feature and removing the page pin allows
several other optimizations that simplify life considerably.
Dependencies:
1) mm_take_all_locks() to register the mmu notifier when the whole VM
isn't doing anything with "mm". This allows mmu notifier users to keep
track if the VM is in the middle of the invalidate_range_begin/end
critical section with an atomic counter incraese in range_begin and
decreased in range_end. No secondary MMU page fault is allowed to map
any spte or secondary tlb reference, while the VM is in the middle of
range_begin/end as any page returned by get_user_pages in that critical
section could later immediately be freed without any further
->invalidate_page notification (invalidate_range_begin/end works on
ranges and ->invalidate_page isn't called immediately before freeing
the page). To stop all page freeing and pagetable overwrites the
mmap_sem must be taken in write mode and all other anon_vma/i_mmap
locks must be taken too.
2) It'd be a waste to add branches in the VM if nobody could possibly
run KVM/GRU/XPMEM on the kernel, so mmu notifiers will only enabled if
CONFIG_KVM=m/y. In the current kernel kvm won't yet take advantage of
mmu notifiers, but this already allows to compile a KVM external module
against a kernel with mmu notifiers enabled and from the next pull from
kvm.git we'll start using them. And GRU/XPMEM will also be able to
continue the development by enabling KVM=m in their config, until they
submit all GRU/XPMEM GPLv2 code to the mainline kernel. Then they can
also enable MMU_NOTIFIERS in the same way KVM does it (even if KVM=n).
This guarantees nobody selects MMU_NOTIFIER=y if KVM and GRU and XPMEM
are all =n.
The mmu_notifier_register call can fail because mm_take_all_locks may be
interrupted by a signal and return -EINTR. Because mmu_notifier_reigster
is used when a driver startup, a failure can be gracefully handled. Here
an example of the change applied to kvm to register the mmu notifiers.
Usually when a driver startups other allocations are required anyway and
-ENOMEM failure paths exists already.
struct kvm *kvm_arch_create_vm(void)
{
struct kvm *kvm = kzalloc(sizeof(struct kvm), GFP_KERNEL);
+ int err;
if (!kvm)
return ERR_PTR(-ENOMEM);
INIT_LIST_HEAD(&kvm->arch.active_mmu_pages);
+ kvm->arch.mmu_notifier.ops = &kvm_mmu_notifier_ops;
+ err = mmu_notifier_register(&kvm->arch.mmu_notifier, current->mm);
+ if (err) {
+ kfree(kvm);
+ return ERR_PTR(err);
+ }
+
return kvm;
}
mmu_notifier_unregister returns void and it's reliable.
The patch also adds a few needed but missing includes that would prevent
kernel to compile after these changes on non-x86 archs (x86 didn't need
them by luck).
[akpm@linux-foundation.org: coding-style fixes]
[akpm@linux-foundation.org: fix mm/filemap_xip.c build]
[akpm@linux-foundation.org: fix mm/mmu_notifier.c build]
Signed-off-by: Andrea Arcangeli <andrea@qumranet.com>
Signed-off-by: Nick Piggin <npiggin@suse.de>
Signed-off-by: Christoph Lameter <cl@linux-foundation.org>
Cc: Jack Steiner <steiner@sgi.com>
Cc: Robin Holt <holt@sgi.com>
Cc: Nick Piggin <npiggin@suse.de>
Cc: Peter Zijlstra <a.p.zijlstra@chello.nl>
Cc: Kanoj Sarcar <kanojsarcar@yahoo.com>
Cc: Roland Dreier <rdreier@cisco.com>
Cc: Steve Wise <swise@opengridcomputing.com>
Cc: Avi Kivity <avi@qumranet.com>
Cc: Hugh Dickins <hugh@veritas.com>
Cc: Rusty Russell <rusty@rustcorp.com.au>
Cc: Anthony Liguori <aliguori@us.ibm.com>
Cc: Chris Wright <chrisw@redhat.com>
Cc: Marcelo Tosatti <marcelo@kvack.org>
Cc: Eric Dumazet <dada1@cosmosbay.com>
Cc: "Paul E. McKenney" <paulmck@us.ibm.com>
Cc: Izik Eidus <izike@qumranet.com>
Cc: Anthony Liguori <aliguori@us.ibm.com>
Cc: Rik van Riel <riel@redhat.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
mm_take_all_locks holds off reclaim from an entire mm_struct. This allows
mmu notifiers to register into the mm at any time with the guarantee that
no mmu operation is in progress on the mm.
This operation locks against the VM for all pte/vma/mm related operations
that could ever happen on a certain mm. This includes vmtruncate,
try_to_unmap, and all page faults.
The caller must take the mmap_sem in write mode before calling
mm_take_all_locks(). The caller isn't allowed to release the mmap_sem
until mm_drop_all_locks() returns.
mmap_sem in write mode is required in order to block all operations that
could modify pagetables and free pages without need of altering the vma
layout (for example populate_range() with nonlinear vmas). It's also
needed in write mode to avoid new anon_vmas to be associated with existing
vmas.
A single task can't take more than one mm_take_all_locks() in a row or it
would deadlock.
mm_take_all_locks() and mm_drop_all_locks are expensive operations that
may have to take thousand of locks.
mm_take_all_locks() can fail if it's interrupted by signals.
When mmu_notifier_register returns, we must be sure that the driver is
notified if some task is in the middle of a vmtruncate for the 'mm' where
the mmu notifier was registered (mmu_notifier_invalidate_range_start/end
is run around the vmtruncation but mmu_notifier_register can run after
mmu_notifier_invalidate_range_start and before
mmu_notifier_invalidate_range_end). Same problem for rmap paths. And
we've to remove page pinning to avoid replicating the tlb_gather logic
inside KVM (and GRU doesn't work well with page pinning regardless of
needing tlb_gather), so without mm_take_all_locks when vmtruncate frees
the page, kvm would have no way to notice that it mapped into sptes a page
that is going into the freelist without a chance of any further
mmu_notifier notification.
[akpm@linux-foundation.org: coding-style fixes]
Signed-off-by: Andrea Arcangeli <andrea@qumranet.com>
Acked-by: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Christoph Lameter <cl@linux-foundation.org>
Cc: Jack Steiner <steiner@sgi.com>
Cc: Robin Holt <holt@sgi.com>
Cc: Nick Piggin <npiggin@suse.de>
Cc: Peter Zijlstra <a.p.zijlstra@chello.nl>
Cc: Kanoj Sarcar <kanojsarcar@yahoo.com>
Cc: Roland Dreier <rdreier@cisco.com>
Cc: Steve Wise <swise@opengridcomputing.com>
Cc: Avi Kivity <avi@qumranet.com>
Cc: Hugh Dickins <hugh@veritas.com>
Cc: Rusty Russell <rusty@rustcorp.com.au>
Cc: Anthony Liguori <aliguori@us.ibm.com>
Cc: Chris Wright <chrisw@redhat.com>
Cc: Marcelo Tosatti <marcelo@kvack.org>
Cc: Eric Dumazet <dada1@cosmosbay.com>
Cc: "Paul E. McKenney" <paulmck@us.ibm.com>
Cc: Izik Eidus <izike@qumranet.com>
Cc: Anthony Liguori <aliguori@us.ibm.com>
Cc: Rik van Riel <riel@redhat.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Li Zefan
Aneesh Kumar K.V
Theodore Ts'o
Eric Sandeen
Hidehiro Kawai
Linus Torvalds
Rusty Russell
Andrew Morton
Mark McLoughlin
Johannes Weiner
Dmitry Baryshkov
Manuel Lauss
Atsushi Nemoto
Yoichi Yuasa
Hisashi Hifumi
Simon Horman
Ben Dooks
Luotao Fu
Adrian Bunk
Andrea Arcangeli