Commit Graph

90 Commits

Author SHA1 Message Date
Aneesh Kumar K.V c7e587b73e hw/9pfs: Fix segfault with 9p2000.u
When guest tries to chmod a block or char device file over 9pfs,
the qemu process segfaults. With 9p2000.u protocol we use wstat to
change mode bits and client don't send extension information for
chmod. We need to check for size field to check whether extension
info is present or not.

Reported-by: Michael Tokarev <mjt@tls.msk.ru>
Acked-by: Michael Tokarev <mjt@tls.msk.ru>
Reviewed-by: Stefan Hajnoczi <stefanha@redhat.com>
Signed-off-by: Aneesh Kumar K.V <aneesh.kumar@linux.vnet.ibm.com>
2013-05-28 15:23:12 +05:30
KONRAD Frederic 13daf6cad0 virtio-9p: cleanup: QOM casts.
As the virtio-9p-pci is switched to the new API, we can use QOM casts.

Signed-off-by: KONRAD Frederic <fred.konrad@greensocs.com>
Message-id: 1366708123-19626-5-git-send-email-fred.konrad@greensocs.com
Signed-off-by: Anthony Liguori <aliguori@us.ibm.com>
2013-04-23 10:37:49 -05:00
Paolo Bonzini 0d09e41a51 hw: move headers to include/
Many of these should be cleaned up with proper qdev-/QOM-ification.
Right now there are many catch-all headers in include/hw/ARCH depending
on cpu.h, and this makes it necessary to compile these files per-target.
However, fixing this does not belong in these patches.

Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
2013-04-08 18:13:10 +02:00
Paolo Bonzini 60653b28f5 virtio-9p: remove PCI dependencies from hw/9pfs/
Also move the 9p.h file to 9pfs/virtio-9p-device.h, for consistency
with the corresponding .c file.

Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
2013-03-01 13:57:33 +01:00
Aneesh Kumar K.V a911a182a6 qemu/9p: Don't ignore error in fid clunk
We use the clunk request to do the actual xattr operation. So don't
ignore the error value for fid clunk.

Security model "none" don't support posix acl. Without this patch
guest won't get EOPNOTSUPP error on setxattr("system.posix_acl_access")

Signed-off-by: Aneesh Kumar K.V <aneesh.kumar@linux.vnet.ibm.com>
Signed-off-by: Anthony Liguori <aliguori@us.ibm.com>
2013-02-06 08:29:50 -06:00
Markus Armbruster 6528499fa4 g_malloc(0) and g_malloc0(0) return NULL; simplify
Once upon a time, it was decided that qemu_malloc(0) should abort.
Switching to glib retired that bright idea.  Some code that was added
to cope with it (e.g. in commits 702ef63, b76b6e9) is still around.
Bury it.

Signed-off-by: Markus Armbruster <armbru@redhat.com>
Reviewed-by: Eric Blake <eblake@redhat.com>
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
2013-01-30 11:14:46 +01:00
Paolo Bonzini 1de7afc984 misc: move include files to include/qemu/
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
2012-12-19 08:32:39 +01:00
Paolo Bonzini caf71f86a3 migration: move include files to include/migration/
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
2012-12-19 08:31:32 +01:00
Amos Kong 4d5b97da35 cleanup useless return sentence
This patch cleans up return sentences in the end of void functions.

Reported-by: Paolo Bonzini <pbonzini@redhat.com>
Signed-off-by: Amos Kong <akong@redhat.com>
Signed-off-by: Stefan Hajnoczi <stefanha@gmail.com>
2012-10-05 15:10:21 +02:00
Aneesh Kumar K.V 4cdc0789ec hw/9pfs: Fix assert when disabling migration
For 9p we can get the attach request multiple times for the
same export. So don't adding migration blocker for every
attach request.

Signed-off-by: Aneesh Kumar K.V <aneesh.kumar@linux.vnet.ibm.com>
2012-07-31 22:01:40 +05:30
Michael Tokarev 1b093c480a consolidate qemu_iovec_copy() and qemu_iovec_concat() and make them consistent
qemu_iovec_concat() is currently a wrapper for
qemu_iovec_copy(), use the former (with extra
"0" arg) in a few places where it is used.

Change skip argument of qemu_iovec_copy() from
uint64_t to size_t, since size of qiov itself
is size_t, so there's no way to skip larger
sizes.  Rename it to soffset, to make it clear
that the offset is applied to src.

Also change the only usage of uint64_t in
hw/9pfs/virtio-9p.c, in v9fs_init_qiov_from_pdu() -
all callers of it actually uses size_t too,
not uint64_t.

One added restriction: as for all other iovec-related
functions, soffset must point inside src.

Order of argumens is already good:
 qemu_iovec_memset(QEMUIOVector *qiov, size_t offset,
                   int c, size_t bytes)
vs:
 qemu_iovec_concat(QEMUIOVector *dst,
                   QEMUIOVector *src,
                   size_t soffset, size_t sbytes)
(note soffset is after _src_ not dst, since it applies to src;
for memset it applies to qiov).

Note that in many places where this function is used,
the previous call is qemu_iovec_reset(), which means
many callers actually want copy (replacing dst content),
not concat.  So we may want to add a wrapper like
qemu_iovec_copy() with the same arguments but which
calls qemu_iovec_reset() before _concat().

Signed-off-by: Michael Tokarev <mjt@tls.msk.ru>
2012-06-11 23:12:11 +04:00
Benjamin Herrenschmidt 67d6fa5362 hw/9pfs: Endian fixes for virtfs
This patch fixes several endian bugs in virtfs.

Cc: Aneesh Kumar K.V <aneesh.kumar@linux.vnet.ibm.com>

Signed-off-by: Benjamin Herrenschmidt <benh@kernel.crashing.org>
Signed-off-by: David Gibson <david@gibson.dropbear.id.au>
Signed-off-by: Aneesh Kumar K.V <aneesh.kumar@linux.vnet.ibm.com>
2012-02-24 14:01:19 +05:30
Daniel P. Berrange eed968607d hw/9pfs: Remove O_NOATIME flag from 9pfs open() calls in readonly mode
When 2c74c2cb4b added support for
the 'readonly' flag against 9p filesystems, it also made QEMU
add the O_NOATIME flag as a side-effect.

The O_NOATIME flag, however, may only be set by the file owner,
or a user with CAP_FOWNER capability.  QEMU cannot assume that
this is the case for filesytems exported to QEMU.

eg, run QEMU as non-root, and attempt to pass the host OS
filesystem through to the guest OS with readonly enable.
The result is that the guest OS cannot open any files at
all.

If O_NOATIME is really required, it should be optionally
enabled via a separate QEMU command line flag.

 * hw/9pfs/virtio-9p.c: Remove O_NOATIME

Acked-by: M. Mohan Kumar <mohan@in.ibm.com>
Signed-off-by: Daniel P. Berrange <berrange@redhat.com>
Signed-off-by: Aneesh Kumar K.V <aneesh.kumar@linux.vnet.ibm.com>
2012-01-30 10:54:17 +05:30
Aneesh Kumar K.V 71f86cd6f3 hw/9pfs: Fix crash when mounting with synthfs
Some Fsdriver backend don't have fs_root. So check for that in
migrate message.

Signed-off-by: Aneesh Kumar K.V <aneesh.kumar@linux.vnet.ibm.com>
2012-01-30 10:54:16 +05:30
M. Mohan Kumar ddca7f86ac hw/9pfs: Add validation to {un}marshal code
Signed-off-by: M. Mohan Kumar <mohan@in.ibm.com>
Signed-off-by: Aneesh Kumar K.V <aneesh.kumar@linux.vnet.ibm.com>
2012-01-04 19:53:22 +05:30
M. Mohan Kumar 10925bf089 hw/9pfs: Move pdu_marshal/unmarshal code to a seperate file
Move p9 marshaling/unmarshaling code to a separate file so that
proxy filesytem driver can use these calls. Also made marshaling
code generic to accept "struct iovec" instead of V9fsPDU.

Signed-off-by: M. Mohan Kumar <mohan@in.ibm.com>
Signed-off-by: Aneesh Kumar K.V <aneesh.kumar@linux.vnet.ibm.com>
2012-01-04 19:51:28 +05:30
Aneesh Kumar K.V e4027caf93 hw/9pfs: iattr_valid flags are kernel internal flags map them to 9p values.
Kernel internal values can change, add protocol values for these constant and
use them.

Signed-off-by: Aneesh Kumar K.V <aneesh.kumar@linux.vnet.ibm.com>
2011-12-21 12:37:23 +05:30
Aneesh Kumar K.V 2f008a8c97 hw/9pfs: Use the correct signed type for different variables
Signed-off-by: Aneesh Kumar K.V <aneesh.kumar@linux.vnet.ibm.com>
2011-12-21 12:37:23 +05:30
Stefan Hajnoczi 302a0d3ed7 hw/9pfs: replace iovec manipulation with QEMUIOVector
The v9fs_read() and v9fs_write() functions rely on iovec[] manipulation
code should be replaced with QEMUIOVector to avoid duplicating code.
In the future it may be possible to make the code even more concise by
using QEMUIOVector consistently across virtio and 9pfs.

The "v" format specifier for pdu_marshal() and pdu_unmarshal() is
dropped since it does not actually pack/unpack anything.  The specifier
was also not implemented to update the offset variable and could only be
used at the end of a format string, another sign that this shouldn't
really be a format specifier.  Instead, see the new
v9fs_init_qiov_from_pdu() function.

This change avoids a possible iovec[] buffer overflow when indirect
vrings are used since the number of vectors is now limited by the
underlying VirtQueueElement and cannot be out-of-bounds.

Signed-off-by: Stefan Hajnoczi <stefanha@linux.vnet.ibm.com>
Signed-off-by: Aneesh Kumar K.V <aneesh.kumar@linux.vnet.ibm.com>
2011-12-21 12:37:22 +05:30
Dong Xu Wang 3a93113a00 fix typo: delete redundant semicolon
Double semicolons should be single.

Signed-off-by: Dong Xu Wang <wdongxu@linux.vnet.ibm.com>
Signed-off-by: Stefan Hajnoczi <stefanha@linux.vnet.ibm.com>
2011-12-06 09:56:41 +00:00
Deepak C Shetty b41e2992b2 hw/9pfs: Reset server state during TVERSION
As per the 9p rfc, during TVERSION its necessary to clean all the active
fids, so that we start the session from a clean state. Its also needed in
scenarios where the guest is booting off 9p, and boot fails, and client
restarts, without any knowledge of the past, it will issue a TVERSION again
so this ensures that we always start from a clean state.

Signed-off-by: Deepak C Shetty <deepakcs@linux.vnet.ibm.com>
Signed-off-by: Aneesh Kumar K.V <aneesh.kumar@linux.vnet.ibm.com>
2011-12-04 22:35:28 +05:30
Aneesh Kumar K.V e9a0152ba1 hw/9pfs: use migration blockers to prevent live migration when virtfs export path is mounted
Now when you try to migrate with VirtFS export path mounted, you get a proper QMP error:

(qemu) migrate tcp:localhost:4444
Migration is disabled when VirtFS export path '/tmp/' is mounted in the guest using mount_tag 'v_tmp'
(qemu)

Signed-off-by: Aneesh Kumar K.V <aneesh.kumar@linux.vnet.ibm.com>
2011-12-04 22:35:28 +05:30
Paolo Bonzini 930b588395 9p: pass dotl flags to the unlinkat method
AT_REMOVEDIR is not defined on all systems.  Pass the raw flags from the
9p protocol, which are always there.

Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
Signed-off-by: Anthony Liguori <aliguori@us.ibm.com>
2011-11-21 14:58:49 -06:00
Stefan Hajnoczi 779b5b04fd hw/9pfs: use g_vasprintf() instead of rolling our own
Markus Armbruster <armbru@redhat.com> sent fixes for va_list vararg
issues in v9fs_string_alloc_printf().  It turns out the function
duplicates g_vasprintf() and can therefore be eliminated entirely.

Signed-off-by: Stefan Hajnoczi <stefanha@linux.vnet.ibm.com>
Signed-off-by: Aneesh Kumar K.V <aneesh.kumar@linux.vnet.ibm.com>
2011-11-02 12:57:15 +05:30
Aneesh Kumar K.V cc720ddb54 hw/9pfs: Abstract open state of fid to V9fsFidOpenState
To implement synthetic file system in Qemu we may not really
require file descriptor and Dir *. Make generic code use
V9fsFidOpenState instead.

Signed-off-by: Aneesh Kumar K.V <aneesh.kumar@linux.vnet.ibm.com>
2011-10-31 12:34:17 +05:30
M. Mohan Kumar 2c74c2cb4b hw/9pfs: Read-only support for 9p export
A new fsdev parameter "readonly" is introduced to control accessing 9p export.
"readonly" can be used to specify the access type. By default "rw" access
is given to 9p export.

Signed-off-by: M. Mohan Kumar <mohan@in.ibm.com>
Signed-off-by: Aneesh Kumar K.V <aneesh.kumar@linux.vnet.ibm.com>
2011-10-31 12:34:17 +05:30
Aneesh Kumar K.V 7999f7e127 hw/9pfs: Make VirtFS tracing work correctly
this patch fix multiple issues with VirtFS tracing.
a) Add tracepoint to the correct code path. We handle error in complete_pdu
b) Fix indentation in python script
c) Fix variable naming issue in python script

Signed-off-by: Aneesh Kumar K.V <aneesh.kumar@linux.vnet.ibm.com>
Signed-off-by: Blue Swirl <blauwirbel@gmail.com>
2011-10-30 09:05:28 +00:00
Aneesh Kumar K.V c98f1d4a8b hw/9pfs: Use export_flag for indicating whether fs driver use path names.
This allows us to remove another member from the struct

Signed-off-by: Aneesh Kumar K.V <aneesh.kumar@linux.vnet.ibm.com>
2011-10-15 15:30:27 +05:30
Harsh Prateek Bora 46604b390e hw/9pfs: Remove virtio-9p-debug.* infra to be replaced by Qemu Tracing.
Removing the existing debug infrastrucure as proposed to be replaced by
Qemu Tracing infrastructure.

Reviewed-by: Stefan Hajnoczi <stefanha@linux.vnet.ibm.com>
Signed-off-by: Harsh Prateek Bora <harsh@linux.vnet.ibm.com>
Signed-off-by: Aneesh Kumar K.V <aneesh.kumar@linux.vnet.ibm.com>
2011-10-15 15:30:27 +05:30
Harsh Prateek Bora c572f23a3e hw/9pfs: Introduce tracing for 9p pdu handlers
Plan is to replace the existing debug infrastructure with Qemu tracing
infrastructure so that user can dynamically enable/disable trace events and
therefore a meaningful trace log can be generated which can be further
filtered using an analysis script.

Note: Because of current simpletrace limitations, the trace events are
logging at max 6 args, however, once the more args are supported, we can
change trace events to log more info as well. Also, This initial patch only
provides a replacement for existing debug infra. More trace events to be
added later for newly added handlers and sub-routines.

Reviewed-by: Stefan Hajnoczi <stefanha@linux.vnet.ibm.com>
Signed-off-by: Harsh Prateek Bora <harsh@linux.vnet.ibm.com>
Signed-off-by: Aneesh Kumar K.V <aneesh.kumar@linux.vnet.ibm.com>
2011-10-15 15:30:27 +05:30
Harsh Prateek Bora e06a765efb hw/9pfs: Add st_gen support in getattr reply
This patch use file system specific ioctl for getting i_generation
value. Not all file system support the ioctl. So we add an export
specific extended operation and assign right callback for the
file system that support i_generation ioctl

["M. Mohan Kumar" <mohan@in.ibm.com> we can do ioctl only for
regular files and directories on the server]

Signed-off-by: Harsh Prateek Bora <harsh@linux.vnet.ibm.com>
Signed-off-by: Aneesh Kumar K.V <aneesh.kumar@linux.vnet.ibm.com>
2011-10-15 15:30:27 +05:30
M. Mohan Kumar 9844081bca hw/9pfs: Add open flag mapping
Some of the flags are OS/arch dependent we need to use
9P defined value on wire,

Based on the original patch from Venkateswararao Jujjuri <jvrao@linux.vnet.ibm.com>

Signed-off-by: M. Mohan Kumar <mohan@in.ibm.com>
Signed-off-by: Aneesh Kumar K.V <aneesh.kumar@linux.vnet.ibm.com>
2011-10-15 15:30:27 +05:30
M. Mohan Kumar ea60f31593 virtio-9p: Use 9P specific Lock constants
Use 9P specific lock constants instead of arch specific lock constants.

Signed-off-by: M. Mohan Kumar <mohan@in.ibm.com>
Signed-off-by: Aneesh Kumar K.V <aneesh.kumar@linux.vnet.ibm.com>
2011-10-15 15:30:26 +05:30
Aneesh Kumar K.V d3ab98e691 hw/9pfs: Add new virtfs option writeout=immediate skip host page cache
writeout=immediate implies the after pwritev we do a sync_file_range.

Reviewed-by: Stefan Hajnoczi <stefanha@linux.vnet.ibm.com>
Signed-off-by: Aneesh Kumar K.V <aneesh.kumar@linux.vnet.ibm.com>
2011-10-12 19:11:23 +05:30
Aneesh Kumar K.V bccacf6c79 hw/9pfs: Implement TFLUSH operation
Signed-off-by: Aneesh Kumar K.V <aneesh.kumar@linux.vnet.ibm.com>
2011-09-22 21:38:52 +05:30
Aneesh Kumar K.V ce421a1961 hw/9pfs: Avoid unnecessary get_fid in v9fs_clunk
Signed-off-by: Aneesh Kumar K.V <aneesh.kumar@linux.vnet.ibm.com>
2011-09-22 21:38:52 +05:30
Aneesh Kumar K.V 532decb715 hw/9pfs: Add fs driver specific details to fscontext
Add a new context flag PATHNAME_FSCONTEXT and indicate whether
the fs driver track fid using path names. Also add a private
pointer that help us to track fs driver specific values in there

Signed-off-by: Aneesh Kumar K.V <aneesh.kumar@linux.vnet.ibm.com>
2011-09-22 21:38:52 +05:30
Aneesh Kumar K.V 2289be19ae hw/9pfs: Move fid pathname tracking to seperate data type.
This enables us to add handles to track fids later. The
V9fsPath added is similar to V9fsString except that the
size include the NULL byte also.

Signed-off-by: Aneesh Kumar K.V <aneesh.kumar@linux.vnet.ibm.com>
2011-09-22 21:38:52 +05:30
Aneesh Kumar K.V 02cb7f3a25 hw/9pfs: Use read-write lock for protecting fid path.
On rename we take the write lock and this ensure path
doesn't change as we operate on them.

Signed-off-by: Aneesh Kumar K.V <aneesh.kumar@linux.vnet.ibm.com>
2011-09-22 21:38:52 +05:30
Aneesh Kumar K.V 70c18fc08b hw/9pfs: Make v9fs_string* functions non-static
We will use them later in other files

Signed-off-by: Aneesh Kumar K.V <aneesh.kumar@linux.vnet.ibm.com>
2011-09-22 21:38:52 +05:30
Aneesh Kumar K.V 7834cf77be hw/9pfs: add 9P2000.L unlinkat operation
unlinkat - Remove a directory entry

size[4] Tunlinkat tag[2] dirfid[4] name[s] flag[4]
size[4] Runlinkat tag[2]

older Tremove have the below request format

size[4] Tremove tag[2] fid[4]

The remove message is used to remove a directory entry either file or directory
The remove opreation is actually a directory opertation and should ideally have
dirfid, if not we cannot represent the fid on server with anything other than
name. We will have to derive the directory name from fid in the Tremove request.

NOTE: The operation doesn't clunk the unlink fid.

Signed-off-by: Aneesh Kumar K.V <aneesh.kumar@linux.vnet.ibm.com>
2011-09-09 15:07:01 +05:30
Aneesh Kumar K.V 89bf65938a hw/9pfs: add 9P2000.L renameat operation
rename - change name of file or directory

size[4] Trenameat tag[2] olddirfid[4] oldname[s] newdirfid[4] newname[s]
size[4] Rrenameat tag[2]

older Trename have the below request format

size[4] Trename tag[2] fid[4] newdirfid[4] name[s]

The rename message is used to change the name of a file, possibly moving it
to a new directory. The rename opreation is actually a directory opertation
and should ideally have olddirfid, if not we cannot represent the fid on server
with anything other than name. We will have to derive the old directory name
from fid in the Trename request.

Signed-off-by: Aneesh Kumar K.V <aneesh.kumar@linux.vnet.ibm.com>
2011-09-09 14:46:03 +05:30
Aneesh Kumar K.V 10e72295bd hw/9pfs: Fix memleaks in some 9p operation
Signed-off-by: Aneesh Kumar K.V <aneesh.kumar@linux.vnet.ibm.com>
2011-09-09 14:46:00 +05:30
Aneesh Kumar K.V 25427ec143 hw/9pfs: Initialize rest of qid field to zero.
Since qid is allocated out of stack we need to intialize
the field to zero. Otherwise we will send wrong qid value
to client.

Signed-off-by: Aneesh Kumar K.V <aneesh.kumar@linux.vnet.ibm.com>
2011-09-09 14:45:55 +05:30
Aneesh Kumar K.V 407fee0f02 hw/9pfs: Update the fidp path before opendir
We need to update the fidp path before opendir. Since we don't
use the fid returned by mkdir, earlier code should not have
much issue. We do a double v9fs_string_copy here. The later patch
cleanup the entire function.

Signed-off-by: Aneesh Kumar K.V <aneesh.kumar@linux.vnet.ibm.com>
2011-09-09 14:45:53 +05:30
Aneesh Kumar K.V b9cb88b056 hw/9pfs: mark directories also as un-reclaimable on unlink
Signed-off-by: Aneesh Kumar K.V <aneesh.kumar@linux.vnet.ibm.com>
2011-08-25 20:16:42 +05:30
Aneesh Kumar K.V 95f65511c3 hw/9pfs: Add directory reclaim support
Signed-off-by: Aneesh Kumar K.V <aneesh.kumar@linux.vnet.ibm.com>
2011-08-25 20:16:42 +05:30
Aneesh Kumar K.V 9b54ffaab2 hw/9pfs: Use v9fs_do_close instead of close
we should use the local abstraction instead of
directly calling close.

Signed-off-by: Aneesh Kumar K.V <aneesh.kumar@linux.vnet.ibm.com>
2011-08-25 20:16:42 +05:30
Aneesh Kumar K.V 7a46274529 hw/9pfs: Add file descriptor reclaim support
[M. Mohan Kumar <mohan@in.ibm.com> removed some unused variables]

Signed-off-by: Aneesh Kumar K.V <aneesh.kumar@linux.vnet.ibm.com>
2011-08-25 20:16:41 +05:30
Aneesh Kumar K.V 84dfb926e5 hw/9pfs: Add reference counting for fid
Signed-off-by: Aneesh Kumar K.V <aneesh.kumar@linux.vnet.ibm.com>
2011-08-25 20:16:41 +05:30