This reverts commit 328bd24443.
As it turns out, this is not portable and very Linux & glibc
specific. Worse, this may lead to not starving writers on Linux
but everywhere else. Revert this and if the starvation occurs
resolve it.
Signed-off-by: Michal Privoznik <mprivozn@redhat.com>
Reviewed-by: Daniel P. Berrange <berrange@redhat.com>
There is no reason why two threads trying to look up two domains
should mutually exclude each other. Utilize new
virObjectRWLockable that was just introduced.
Signed-off-by: Michal Privoznik <mprivozn@redhat.com>
Reviewed-by: Pavel Hrdina <phrdina@redhat.com>
Up until now we only had virObjectLockable which uses mutexes for
mutually excluding each other in critical section. Well, this is
not enough. Future work will require RW locks so we might as well
have virObjectRWLockable which is introduced here.
Moreover, polymorphism is introduced to our code for the first
time. Yay! More specifically, virObjectLock will grab a write
lock, virObjectLockRead will grab a read lock then (what a
surprise right?). This has great advantage that an object can be
made derived from virObjectRWLockable in a single line and still
continue functioning properly (mutexes can be viewed as grabbing
write locks only). Then just those critical sections that can
grab a read lock need fixing. Therefore the resulting change is
going to be way smaller.
In order to avoid writer starvation, the object initializes RW
lock that prefers writers.
Signed-off-by: Michal Privoznik <mprivozn@redhat.com>
Reviewed-by: Pavel Hrdina <phrdina@redhat.com>
We already have virRWLockInit. But this uses pthread defaults
which prefer reader to initialize the RW lock. This may lead to
writer starvation. Therefore we need to have the counterpart that
prefers writers. Now, according to the
pthread_rwlockattr_setkind_np() man page setting
PTHREAD_RWLOCK_PREFER_WRITER_NP attribute is no-op. Therefore we
need to use PTHREAD_RWLOCK_PREFER_WRITER_NONRECURSIVE_NP
attribute. So much for good enum value names.
Signed-off-by: Michal Privoznik <mprivozn@redhat.com>
Reviewed-by: Pavel Hrdina <phrdina@redhat.com>
When generating HTML documentation we put enum values into a
table so that we can display the value's name, numerical value
and description (if it has one). Now the last part is problem. If
the value doesn't have description the table row has just two
cells and if it has one the row counts three cells. This makes
HTML engines render the description into very little space - for
instance see:
html/libvirt-libvirt-domain.html#virDomainMemoryStatTags
We can avoid this problem if we let the cell that corresponds to
numerical value span over two cells if there's no description.
Signed-off-by: Michal Privoznik <mprivozn@redhat.com>
Reviewed-by: Martin Kletzander <mkletzan@redhat.com>
After f4cb85c6af we only have two options for placing enum
values descriptions. It's either:
typedef enum {
/* Some long description. Therefore it's placed before
* the value. */
VIR_ENUM_A_VAL = 1,
} virEnumA;
or:
typedef enum {
VIR_ENUM_B_VAL = 1, /* Some short description */
} virEnumB;
However, our apibuild.py script is not able to deal with the
former one. It messes up comments. To fix this couple of things
needs to be done:
a) DO NOT reset self.comment in parseEnumBlock(). This is a
result from our tokenizer. Upon calling token() if it finds a
comment block it stores it in self.comment and returns the next
token (which is not comment). Therefore, if we reset self.comment
we might lose the first comment in the enum block.
b) we need a variable to track if the current enum block uses
value descriptions before or after values. That is if it's type
virEnumA or virEnumB. Depending on that, it we're dealing with
virEnumA type and the current token is a comma ',' we can add the
value into the list as we already have everything needed:
comment, name and value.
Signed-off-by: Michal Privoznik <mprivozn@redhat.com>
Reviewed-by: Martin Kletzander <mkletzan@redhat.com>
So the way we format this huge virQEMUCaps enum is we group the
values in groups of five. And then at the beginning of each group
we have a small comment that says what's the number of the first
item in the group. Well, the last commit of 11b2ebf3e1 does not
follow this formatting.
Signed-off-by: Michal Privoznik <mprivozn@redhat.com>
There are only two acceptable places for describing enum values.
It's either:
typedef enum {
/* Some long description. Therefore it's placed before
* the value. */
VIR_ENUM_A_VAL = 1,
} virEnumA;
or:
typedef enum {
VIR_ENUM_B_VAL = 1, /* Some short description */
} virEnumB;
However, during review of a patch sent upstream I realized that
is not always the case. I went through all the public header
files and identified all the offenders. Luckily there were just
two of them.
Yes, this makes our HTML generated documentation broken, but
that's bug of the generator. Our header files shouldn't be forced
to use something we don't want to.
Signed-off-by: Michal Privoznik <mprivozn@redhat.com>
The purpose of this function is to tell if the current position
in given FD is in data section or a hole and how much bytes there
is remaining until the end of the section. This is achieved by
couple of lseeks(). The most important part is that we reposition
the FD back, so that the position is unchanged from the caller
POV. And until now the final lseek() back to the original
position was done with no check for errors. And I was convinced
that that's okay since nothing can go wrong. However, review
feedback from a related series persuaded me, that it's better to
be safe than sorry. Therefore, lets check if the final lseek()
succeeded and if it doesn't report an error.
Signed-off-by: Michal Privoznik <mprivozn@redhat.com>
Reviewed-by: John Ferlan <jferlan@redhat.com>
In Wireshark commit of 7cd6906056922e4b8 (contained in v2.4.0)
the tvb_new_subset() function was renamed to
tvb_new_subset_length_caplen(). However, we can take the extra
step and rename to tvb_new_subset_remaining() directly (see
Wireshark commit 0ecfc7280cf3d7). The reasoning is that there is
no other protocol in the packet than libvirt. Therefore, from the
point that libvirt dissector takes over till the end of the
packet it's all libvirt packet.
Signed-off-by: Michal Privoznik <mprivozn@redhat.com>
Reviewed-by: Martin Kletzander <mkletzan@redhat.com>
Obviously, old gcc-s ale sad when a variable shares the name with
a function. And we do have such variable (added in 4d8a914be0):
@mount. Rename it to @mountpoint so that compiler's happy again.
Signed-off-by: Michal Privoznik <mprivozn@redhat.com>
The way we create devices under /dev is highly linux specific.
For instance we do mknod(), mount(), umount(), etc. Some
platforms are even missing some of these functions. Then again,
as declared in qemuDomainNamespaceAvailable(): namespaces are
linux only. Therefore, to avoid obfuscating the code by trying to
make it compile on weird platforms, just provide a non-linux stub
for qemuDomainAttachDeviceMknodRecursive(). At the same time,
qemuDomainAttachDeviceMknodHelper() which actually calls the
non-existent functions is moved under ifdef __linux__ block since
its only caller is in that block too.
Signed-off-by: Michal Privoznik <mprivozn@redhat.com>
Reviewed-by: Daniel P. Berrange <berrange@redhat.com>
Currently, the only type of chardev that we create the backend
for in the namespace is type='dev'. This is not enough, other
backends might have files under /dev too. For instance channels
might have a unix socket under /dev (well, bind mounted under
/dev from a different place).
Signed-off-by: Michal Privoznik <mprivozn@redhat.com>
Reviewed-by: John Ferlan <jferlan@redhat.com>
https://bugzilla.redhat.com/show_bug.cgi?id=1462060
Just like in the previous commit, when attaching a file based
device which has its source living under /dev (that is not a
device rather than a regular file), calling mknod() is no help.
We need to:
1) bind mount device to some temporary location
2) enter the namespace
3) move the mount point to desired place
4) umount it in the parent namespace from the temporary location
At the same time, the check in qemuDomainNamespaceSetupDisk makes
no longer sense. Therefore remove it.
Signed-off-by: Michal Privoznik <mprivozn@redhat.com>
Reviewed-by: John Ferlan <jferlan@redhat.com>
https://bugzilla.redhat.com/show_bug.cgi?id=1462060
When building a qemu namespace we might be dealing with bare
regular files. Files that live under /dev. For instance
/dev/my_awesome_disk:
<disk type='file' device='disk'>
<driver name='qemu' type='qcow2'/>
<source file='/dev/my_awesome_disk'/>
<target dev='vdc' bus='virtio'/>
</disk>
# qemu-img create -f qcow2 /dev/my_awesome_disk 10M
So far we were mknod()-ing them which is
obviously wrong. We need to touch the file and bind mount it to
the original:
1) touch /var/run/libvirt/qemu/fedora.dev/my_awesome_disk
2) mount --bind /dev/my_awesome_disk /var/run/libvirt/qemu/fedora.dev/my_awesome_disk
Later, when the new /dev is built and replaces original /dev the
file is going to live at expected location.
Signed-off-by: Michal Privoznik <mprivozn@redhat.com>
Reviewed-by: John Ferlan <jferlan@redhat.com>
Currently, we silently assume that file we are creating in the
namespace is either a link or a device (character or block one).
This is not always the case. Therefore instead of doing something
wrong, claim about unsupported file type.
Signed-off-by: Michal Privoznik <mprivozn@redhat.com>
Reviewed-by: John Ferlan <jferlan@redhat.com>
Currently, we silently assume that file we are creating in the
namespace is either a link or a device (character or block one).
This is not always the case. Therefore instead of doing something
wrong, claim about unsupported file type.
Signed-off-by: Michal Privoznik <mprivozn@redhat.com>
Reviewed-by: John Ferlan <jferlan@redhat.com>
This function is going to be used on other places, so
instead of copying code we can just call the function.
Signed-off-by: Michal Privoznik <mprivozn@redhat.com>
Reviewed-by: John Ferlan <jferlan@redhat.com>
https://bugzilla.redhat.com/show_bug.cgi?id=1459592
In 290a00e41d I've tried to fix the process of building a
qemu namespace when dealing with file mount points. What I
haven't realized then is that we might be dealing not with just
regular files but also special files (like sockets). Indeed, try
the following:
1) socat unix-listen:/tmp/soket stdio
2) touch /dev/socket
3) mount --bind /tmp/socket /dev/socket
4) virsh start anyDomain
Problem with my previous approach is that I wasn't creating the
temporary location (where mount points under /dev are moved) for
anything but directories and regular files.
Signed-off-by: Michal Privoznik <mprivozn@redhat.com>
Reviewed-by: John Ferlan <jferlan@redhat.com>
It comes very handy to have source path for chardevs. We already
have such function: virDomainAuditChardevPath() but it's static
and has name not suitable for exposing. Moreover, while exposing
it change its name slightly to virDomainChrSourceDefGetPath.
Signed-off-by: Michal Privoznik <mprivozn@redhat.com>
Reviewed-by: John Ferlan <jferlan@redhat.com>
All of these four functions (virStreamRecvAll, virStreamSendAll,
virStreamSparseRecvAll, virStreamSparseSendAll) take one or more
callback functions that handle various aspects of streams.
However, if any of them fails no error is reported therefore
caller does not know what went wrong.
At the same time, we silently presumed callbacks to set errno on
failure. With this change we should document it explicitly as the
error is not properly reported.
Signed-off-by: Michal Privoznik <mprivozn@redhat.com>
Reviewed-by: John Ferlan <jferlan@redhat.com>
If one these four functions fail (virStreamRecvAll,
virStreamSendAll, virStreamSparseRecvAll, virStreamSparseSendAll)
the stream is aborted by calling virStreamAbort(). This is a
public API; therefore, the first thing it does is error reset. At
that point any error that caused us to abort stream in the first
place is gone.
Signed-off-by: Michal Privoznik <mprivozn@redhat.com>
Reviewed-by: John Ferlan <jferlan@redhat.com>
Our documentation to the virStreamRecvAll, virStreamSendAll,
virStreamSparseRecvAll, and virStreamSparseSendAll functions
indicates that if these functions fail, then virStreamAbort is
called. But that is not necessarily true. For instance all of
these functions allocate a buffer to work with. If the allocation
fails, no virStreamAbort() is called despite -1 being returned.
It's the same story with argument sanity checks and a lot of
other checks.
Signed-off-by: Michal Privoznik <mprivozn@redhat.com>
Reviewed-by: John Ferlan <jferlan@redhat.com>
Problem with our error reporting is that the error object is a
thread local variable. That means if there's an error reported
within the I/O thread it gets logged and everything, but later
when the event loop aborts the stream it doesn't see the original
error. So we are left with some generic error. We can do better
if we copy the error message between the threads.
Signed-off-by: Michal Privoznik <mprivozn@redhat.com>
Reviewed-by: John Ferlan <jferlan@redhat.com>
When the I/O thread quits (e.g. due to an I/O error, lseek()
error, whatever), any subsequent virFDStream API should return
error too. Moreover, when invoking stream event callback, we must
set the VIR_STREAM_EVENT_ERROR flag so that the callback knows
something bad happened.
Signed-off-by: Michal Privoznik <mprivozn@redhat.com>
Reviewed-by: John Ferlan <jferlan@redhat.com>
Since qemu commit 3ef6c40ad0b it can fail if trying to hotplug a
disk that is not qcow2 despite us saying it is. We need to error
out in that case.
Signed-off-by: Michal Privoznik <mprivozn@redhat.com>
Reviewed-by: John Ferlan <jferlan@redhat.com>
https://bugzilla.redhat.com/show_bug.cgi?id=1461270
When fetching stats for a vhost-user type of interface, we run
couple of ovs-vsctl commands and parse their output. However, not
all stats exist at all times, for instance "rx_dropped" or
"tx_errors" can be missing. Thing is, we ask for a bulk of
statistics and if one of them is missing an error is reported
instead of returning the rest. Since we ignore errors, we fail to
set statistics. Fix this by asking for each piece alone.
Signed-off-by: Michal Privoznik <mprivozn@redhat.com>
Our commit e13e8808f9 was way too generic. Currently, virtlogd is
used only for chardevs type of file and nothing else. True, we
must not relabel the path in this case, but we have to in all
other cases. For instance, if you want to have a physical console
attached to your guest:
<console type='dev'>
<source path='/dev/ttyS0'/>
<target type='virtio' port='1'/>
</console>
Starting such domain fails because qemu doesn't have access to
/dev/ttyS0 because we haven't relabelled the path.
Signed-off-by: Michal Privoznik <mprivozn@redhat.com>
Reviewed-by: John Ferlan <jferlan@redhat.com>
This function is called unconditionally from qemuProcessStop to
make sure we leave no dangling dirs behind. However, whenever the
directory we want to rmdir() is not there (e.g. because it hasn't
been created in the first place because domain doesn't use
hugepages at all), we produce a warning like this:
2017-06-20 15:58:23.615+0000: 32638: warning :
qemuProcessBuildDestroyHugepagesPath:3363 : Unable to remove
hugepage path: /dev/hugepages/libvirt/qemu/1-instance-00000001
(errno=2)
Fix this by not producing the warning on ENOENT.
Signed-off-by: Michal Privoznik <mprivozn@redhat.com>
Reviewed-by: Martin Kletzander <mkletzan@redhat.com>
https://bugzilla.redhat.com/show_bug.cgi?id=1431112
Imagine a FS mounted on /dev/blah/blah2. Our process of creating
suffix for temporary location where all the mounted filesystems
are moved is very simplistic. We want:
/var/run/libvirt/qemu/$domName.$suffix\
were $suffix is just the mount point path stripped of the "/dev/"
prefix. For instance:
/var/run/libvirt/qemu/fedora.mqueue for /dev/mqueue
/var/run/libvirt/qemu/fedora.pts for /dev/pts
and so on. Now if we plug /dev/blah/blah2 into the example we see
some misbehaviour:
/var/run/libvirt/qemu/fedora.blah/blah2
Well, misbehaviour if /dev/blah/blah2 is a file, because in that
case we call virFileTouch() instead of virFileMakePath().
The solution is to replace all the slashes in the suffix with say
dots. That way we don't have to care about nested directories.
IOW, the result we want for given example is:
/var/run/libvirt/qemu/fedora.blah.blah2
Signed-off-by: Michal Privoznik <mprivozn@redhat.com>
Reviewed-by: John Ferlan <jferlan@redhat.com>
https://bugzilla.redhat.com/show_bug.cgi?id=1431112
There can be nested mount points. For instance /dev/shm/blah can
be a mount point and /dev/shm too. It doesn't make much sense to
return the former path because callers preserve the latter (and
with that the former too). Therefore prune nested mount points.
Signed-off-by: Michal Privoznik <mprivozn@redhat.com>
Reviewed-by: John Ferlan <jferlan@redhat.com>
https://bugzilla.redhat.com/show_bug.cgi?id=1431112
After 290a00e41d we know how to deal with file mount points.
However, when cleaning up the temporary location for preserved
mount points we are still calling rmdir(). This won't fly for
files. We need to call unlink(). Now, since we don't really care
if the cleanup succeeded or not (it's the best effort anyway), we
can call both rmdir() and unlink() without need for
differentiation between files and directories.
Signed-off-by: Michal Privoznik <mprivozn@redhat.com>
Reviewed-by: John Ferlan <jferlan@redhat.com>
https://bugzilla.redhat.com/show_bug.cgi?id=1214369
My fix 671d18594f was incomplete. If domain doesn't have
hugepages enabled, because of missing condition we would still be
putting hugepages path onto qemu cmd line. Clean up the
conditions so that it's more visible next time.
Signed-off-by: Michal Privoznik <mprivozn@redhat.com>
https://bugzilla.redhat.com/show_bug.cgi?id=1214369
Consider the following XML:
<memoryBacking>
<hugepages>
<page size='2048' unit='KiB' nodeset='1'/>
</hugepages>
<source type='file'/>
<access mode='shared'/>
</memoryBacking>
<numa>
<cell id='0' cpus='0-3' memory='512000' unit='KiB'/>
<cell id='1' cpus='4-7' memory='512000' unit='KiB'/>
</numa>
The following cmd line is generated:
-object
memory-backend-file,id=ram-node0,mem-path=/var/lib/libvirt/qemu/ram,
share=yes,size=524288000 -numa node,nodeid=0,cpus=0-3,memdev=ram-node0
-object
memory-backend-file,id=ram-node1,mem-path=/var/lib/libvirt/qemu/ram,
share=yes,size=524288000 -numa node,nodeid=1,cpus=4-7,memdev=ram-node1
This is obviously wrong as for node 1 hugepages should have been
used. The hugepages configuration is more specific than <source
type='file'/>.
Signed-off-by: Michal Privoznik <mprivozn@redhat.com>
We have couple of hugepage enabled domains for qemuxml2argvtest.
Unfortunately, often when adding a test case there I forget to
add it to xml2xml test too.
Signed-off-by: Michal Privoznik <mprivozn@redhat.com>
https://bugzilla.redhat.com/show_bug.cgi?id=1455819
It may happen that a domain is started without any huge pages.
However, user might try to attach a DIMM module later. DIMM
backed by huge pages (why would somebody want to mix regular and
huge pages is beyond me). Therefore we have to create the dir if
we haven't done so far.
Signed-off-by: Michal Privoznik <mprivozn@redhat.com>
Reviewed-by: John Ferlan <jferlan@redhat.com>
https://bugzilla.redhat.com/show_bug.cgi?id=1455819
Currently, the per-domain path for huge pages mmap() for qemu is
created iff domain has memoryBacking and hugepages in it
configured. However, this alone is not enough because there can
be a DIMM module with hugepages configured too.
Signed-off-by: Michal Privoznik <mprivozn@redhat.com>
Reviewed-by: John Ferlan <jferlan@redhat.com>
Right now, there is a lot of exit points from the function.
Depending on their position they need to copy the same free
calls. This goes against our style where we usually have just one
exit point from the function which also does the necessary free.
Signed-off-by: Michal Privoznik <mprivozn@redhat.com>
https://bugzilla.redhat.com/show_bug.cgi?id=1447618
Currently, any attempt to change MTU on an interface that is
plugged to a running domain is silently ignored. We should either
do what's asked or error out. Well, we can update the host side
of the interface, but we cannot change 'host_mtu' attribute for
the virtio-net device. Therefore we have to error out.
Signed-off-by: Michal Privoznik <mprivozn@redhat.com>
Reviewed-by: Laine Stump <laine@laine.org>
https://bugzilla.redhat.com/show_bug.cgi?id=1408701
While implementing MTU (572eda12ad and friends), I've forgotten
to actually set MTU on the host NIC in case of hotplug. We
correctly tell qemu on the monitor what the MTU should be, but we
are not actually setting it on the host NIC.
Signed-off-by: Michal Privoznik <mprivozn@redhat.com>
Reviewed-by: Laine Stump <laine@laine.org>
https://bugzilla.redhat.com/show_bug.cgi?id=1459091
We try to get the last element of the passed path by calling
strrch(path, '/'). However, the pointer that strrchr() returns
points at the slash, We want string that starts right after that.
Signed-off-by: Michal Privoznik <mprivozn@redhat.com>
https://bugzilla.redhat.com/show_bug.cgi?id=1459091
Currently, we are querying for vhostuser interface name in post
parse callback. At that time interface might not yet exist.
However, it has to exist when starting domain. Therefore it makes
more sense to query its name at that point. This partially
reverts 57b5e27.
Signed-off-by: Michal Privoznik <mprivozn@redhat.com>
While reworking client side of streams, I had to postpone payload
decoding so that stream holes and stream data can be
distinguished in virNetClientStreamRecvPacket. That's merely what
18944b7aea does. However, I accidentally removed one important
bit: when server sends us an empty STREAM packet (with no
payload) - meaning end of stream - st->incomingEOF flag needs to
be set. It used to be before I touched the code. After I removed
it, virNetClientStreamRecvPacket will try to fetch more data from
the stream, but it will never come.
Signed-off-by: Michal Privoznik <mprivozn@redhat.com>
Reviewed-by: Jim Fehlig <jfehlig@suse.com>
In 9cb891141c we've introduced some logic to clearing suggested
macvtap/macvlan ifnames. The logic consists of comparing ifname
string with strings that libvirt would generate. However, due to
a typo only VIR_NET_GENERATED_MACVTAP_PREFIX was compared. Twice.
Signed-off-by: Michal Privoznik <mprivozn@redhat.com>
There's a problem with current streams after I switched them from
iohelper to thread implementation. Previously, iohelper made sure
not to exceed specified @length resulting in the pipe EOF
appearing at the exact right moment (the pipe was used to tunnel
the data from the iohelper to the daemon). Anyway, when switching
to thread I had to write the I/O code from scratch. Whilst doing
that I took an inspiration from the iohelper code, but since the
usage of pipe switched to slightly different meaning, there was
no 1:1 relationship between the codes.
Moreover, after introducing VIR_FDSTREAM_MSG_TYPE_HOLE, the
condition that should made sure we won't exceed @length was
completely wrong.
The fix is to:
a) account for holes for @length
b) cap not just data sections but holes too (if @length would be
exceeded)
For this purpose, the condition needs to be brought closer to the
code that handles holes and data sections.
Signed-off-by: Michal Privoznik <mprivozn@redhat.com>
In 4f0aeed I've expanded the list of arguments for
virDomainDefCheckABIStability() but I forgot to fix
bhyveargv2xmltest.
Signed-off-by: Michal Privoznik <mprivozn@redhat.com>
While checking for ABI stability, drivers might pose additional
checks that are not valid for general case. For instance, qemu
driver might check some memory backing attributes because of how
qemu works. But those attributes may work well in other drivers.
Signed-off-by: Michal Privoznik <mprivozn@redhat.com>
In 48d9e6cdcc and friends we've allowed users to back guest
memory by a file inside the host. And in order to keep things
manageable the memory_backing_dir variable was introduced to
qemu.conf to specify the directory where the files are kept.
However, libvirt's policy is that directories are created on
domain startup if they don't exist. We've missed this one.
Signed-off-by: Michal Privoznik <mprivozn@redhat.com>
The @cpus is allocated by virFileReadValueBitmap() but never
freed:
==21274== 40 (32 direct, 8 indirect) bytes in 1 blocks are definitely lost in loss record 808 of 1,004
==21274== at 0x4C2E080: calloc (vg_replace_malloc.c:711)
==21274== by 0x54BA561: virAlloc (viralloc.c:144)
==21274== by 0x54BC604: virBitmapNewEmpty (virbitmap.c:126)
==21274== by 0x54BD059: virBitmapParseUnlimited (virbitmap.c:570)
==21274== by 0x54EECE9: virFileReadValueBitmap (virfile.c:4113)
==21274== by 0x5563132: virCapabilitiesInitCaches (capabilities.c:1548)
==21274== by 0x2BB86E59: virQEMUCapsInit (qemu_capabilities.c:1132)
==21274== by 0x2BBEC067: virQEMUDriverCreateCapabilities (qemu_conf.c:928)
==21274== by 0x2BC3DEAA: qemuStateInitialize (qemu_driver.c:845)
==21274== by 0x5625AAC: virStateInitialize (libvirt.c:770)
==21274== by 0x124519: daemonRunStateInit (libvirtd.c:881)
==21274== by 0x554C927: virThreadHelper (virthread.c:206)
Signed-off-by: Michal Privoznik <mprivozn@redhat.com>
There's a slight problem with the current function. Assume we are
currently in a data section and we have say 42 bytes until next
section. Therefore, just before (handler) is called to fill up
the buffer with data, @want is changed to 42 to match the amount
of data left in the current section. However, after hole is
processed, we are back in data section but with incredibly small
@want size. Nobody will ever reset it back. This results in
incredible data fragmentation.
Signed-off-by: Michal Privoznik <mprivozn@redhat.com>
To every virDirOpen we must have VIR_DIR_CLOSE otherwise FD is
leaked.
Signed-off-by: Michal Privoznik <mprivozn@redhat.com>
Reviewed-by: Pavel Hrdina <phrdina@redhat.com>
On systems with older glibc including fcntl.h for getting
FALLOC_FL_PUNCH_HOLE defined is not enough. We must also include
linux/falloc.h.
Signed-off-by: Michal Privoznik <mprivozn@redhat.com>
Some older systems (such as RHEL6) lack SEEK_HOLE and SEEK_DATA
which virFileInData relies on. Provide a stub for these systems.
Signed-off-by: Michal Privoznik <mprivozn@redhat.com>
Currently, we don't assign any meaning to that. Our current view
on virStream is that it's merely a pipe. And pipes don't support
seeking.
Signed-off-by: Michal Privoznik <mprivozn@redhat.com>
Similarly to previous commit, implement sparse streams feature
for vol-upload. This is, however, slightly different approach,
because we must implement a function that will tell us whether
we are in a data section or in a hole. But there's no magic
hidden in here.
Signed-off-by: Michal Privoznik <mprivozn@redhat.com>
Add a new --sparse switch that does nothing more than
enables the sparse streams feature for this command. Among with
the switch new helper function is introduced: virshStreamSkip().
This is the callback that is called whenever daemon sends us a
hole. In the callback we reflect the hole in underlying file by
seeking as many bytes as told.
Signed-off-by: Michal Privoznik <mprivozn@redhat.com>
These flags to APIs will tell if caller wants to use sparse
stream for storage transfer. At the same time, it's safe to
enable them in storage driver frontend and rely on our backends
checking the flags. This way we can enable specific flags only on
some specific backends, e.g. enable
VIR_STORAGE_VOL_DOWNLOAD_SPARSE_STREAM for filesystem backend but
not iSCSI backend.
Signed-off-by: Michal Privoznik <mprivozn@redhat.com>
Now, not all APIs are going to support sparse streams. To some it
makes no sense at all, e.g. virDomainOpenConsole() or
virDomainOpenChannel(). To others, we will need a special flag to
indicate that client wants to enable sparse streams. Instead of
having to write RPC dispatchers by hand we can just annotate in
our .x files that a certain flag to certain RPC call enables this
feature. For instance:
/**
* @generate: both
* @readstream: 1
* @sparseflag: VIR_SPARSE_STREAM
* @acl: storage_vol:data_read
*/
REMOTE_PROC_DOMAIN_SOME_API = XXX,
Therefore, whenever client calls virDomainSomeAPI(..,
VIR_SPARSE_STREAM); daemon will mark that down and send stream
skips when possible.
Signed-off-by: Michal Privoznik <mprivozn@redhat.com>
Basically, what is needed here is to introduce new message type
for the messages passed between the event loop callbacks and the
worker thread that does all the I/O. The idea is that instead of
a queue of read buffers we will have a queue where "hole of size
X" messages appear. That way the event loop callbacks can just
check the head of the queue and see if the worker thread is in
data or a hole section and how long the section is.
Signed-off-by: Michal Privoznik <mprivozn@redhat.com>
Whenever client is able to receive some data from stream
daemonStreamHandleRead is called. But now the behaviour of this
function needs to be changed a bit. Previously it just read data
from underlying file (of chardev or whatever) and sent those
through the stream to client. This model will not work any longer
because it does not differentiate whether underlying file is in
data or hole section. Therefore, at the beginning of this
function add code that checks this situation and acts
accordingly.
So after the this, when wanting to send some data we always check
whether we are not in a hole and if so, skip it an inform client
about its size.
Signed-off-by: Michal Privoznik <mprivozn@redhat.com>
Whenever server sends a client stream packet (either regular with
actual data or stream skip one) it is queued on @st->rx. So the
list is a mixture of both types of stream packets. So now that we
have all the helpers needed we can wire their processing up. But
since virNetClientStreamRecvPacket doesn't support
VIR_STREAM_RECV_STOP_AT_HOLE flag yet, let's turn all received
skips into zeroes repeating requested times.
Signed-off-by: Michal Privoznik <mprivozn@redhat.com>
Now that we have RPC wrappers over VIR_NET_STREAM_HOLE we can
start wiring them up. This commit wires up situation when a
client wants to send a hole to daemon.
To keep stream offsets synchronous, upon successful call on the
daemon skip the same hole in local part of the stream.
Signed-off-by: Michal Privoznik <mprivozn@redhat.com>
This is a function that handles an incoming STREAM_HOLE packet.
Even though it is not wired up yet, it will be soon. At the
beginning do couple of checks whether server plays nicely and
sent us a STREAM_HOLE packed only after we've enabled sparse
streams. Then decodes the message payload to see how big the hole
is and stores it in passed @length argument.
Signed-off-by: Michal Privoznik <mprivozn@redhat.com>
Basically, whenever the new type of stream packet arrives to the
daemon call this function that decodes it and calls
virStreamSendHole(). Otherwise a regular data stream packet has
arrived and therefore continue its processing.
Signed-off-by: Michal Privoznik <mprivozn@redhat.com>
While the previous commit implemented a helper for sending a
STREAM_HOLE packet for daemon, this is a client's counterpart.
Signed-off-by: Michal Privoznik <mprivozn@redhat.com>
This is just a helper function that takes in a length value,
encodes it into XDR and sends to client.
Signed-off-by: Michal Privoznik <mprivozn@redhat.com>
Ideally, this would be generated, but to achieve that
corresponding XDR definitions needed to go into a different .x
file. But they belong just to the one that they are right now.
Signed-off-by: Michal Privoznik <mprivozn@redhat.com>
This is a special type of stream packet, that is bidirectional
and contains information regarding how many bytes each side will
be skipping in the stream.
Signed-off-by: Michal Privoznik <mprivozn@redhat.com>
Add a new argument to daemonCreateClientStream in order to allow for
future expansion to mark that a specific stream can be used to skip
data, such as the case with sparsely populated files. The new flag will
be the eventual decision point between client/server to decide whether
both ends can support and want to use sparse streams.
A new bool 'allowSkip' is added to both _virNetClientStream and
daemonClientStream in order to perform the tracking.
Signed-off-by: Michal Privoznik <mprivozn@redhat.com>
Add a virStreamPtr pointer to the _virNetClientStream
in order to reverse track the parent stream.
Signed-off-by: Michal Privoznik <mprivozn@redhat.com>
This is just an internal API, that calls corresponding function
in stream driver. This function will set @data = 1 if the
underlying file is in data section, or @data = 0 if it is in a
hole. At any rate, @length is set to number of bytes remaining in
the section the file currently is.
Signed-off-by: Michal Privoznik <mprivozn@redhat.com>
This is just a wrapper over new function that have been just
introduced: virStreamSendHole() . It's very similar to
virStreamSendAll() except it handles sparse streams well.
Signed-off-by: Michal Privoznik <mprivozn@redhat.com>
This is just a wrapper over new functions that have been just
introduced: virStreamRecvFlags(), virStreamRecvHole(). It's very
similar to virStreamRecvAll() except it handles sparse streams
well.
Signed-off-by: Michal Privoznik <mprivozn@redhat.com>
Add a new flag to virStreamRecvFlags in order to handle being able to
stop reading from the stream so that the consumer can generate a "hole"
in stream target. Generation of a hole replaces the need to receive and
handle a sequence of zero bytes for sparse stream targets.
Signed-off-by: Michal Privoznik <mprivozn@redhat.com>
This function is basically a counterpart for virStreamSendHole().
If one side of a stream called virStreamSendHole() the other
should call virStreamRecvHole() to get the size of the hole.
Signed-off-by: Michal Privoznik <mprivozn@redhat.com>
This API is used to tell the other side of the stream to skip
some bytes in the stream. This can be used to create a sparse
file on the receiving side of a stream.
It takes @length argument, which says how big the hole is. This
skipping is done from the current point of stream. Since our
streams are not rewindable like regular files, we don't need
@whence argument like seek(2) has.
Signed-off-by: Michal Privoznik <mprivozn@redhat.com>
There are three virStreamDriver's currently supported:
* virFDStream
* remote driver
* ESX driver
For now, backend virStreamRecvFlags support for only remote driver and
ESX driver is sufficient. Future patches will update virFDStream.
Signed-off-by: Michal Privoznik <mprivozn@redhat.com>
This patch is adding the virStreamRecvFlags as a variant to the
virStreamRecv function in order to allow for future expansion of
functionality for processing sparse streams using a @flags
argument.
Signed-off-by: Michal Privoznik <mprivozn@redhat.com>
This function takes a FD and determines whether the current
position is in data section or in a hole. In addition to that,
it also determines how much bytes are there remaining till the
current section ends.
Signed-off-by: Michal Privoznik <mprivozn@redhat.com>
One big downside of using the pipe to transfer the data is that
we can really transfer just bare data. No metadata can be carried
through unless some formatted messages are introduced. That would
be quite painful to achieve so let's use a message queue. It's
fairly easy to exchange info between threads now that iohelper is
no longer used.
The reason why we cannot use the FD for plain files directly is
that despite us setting noblock flag on the FD, any
read()/write() blocks regardless (which is a show stopper since
those parts of the code are run from the event loop) and poll()
reports such FD as always readable/writable - even though the
subsequent operation might block.
The pipe is still not gone though. It is used to signal the event
loop that an event occurred (e.g. data is available for reading
in the queue, or vice versa).
Signed-off-by: Michal Privoznik <mprivozn@redhat.com>
When adding a nwfilter onto the list in
virNWFilterObjListAssignDef() this array is re-allocated to match
demand for new size. However, it is never freed leading to a
leak:
==26535== 136 bytes in 1 blocks are definitely lost in loss record 1,079 of 1,250
==26535== at 0x4C2E2BE: realloc (vg_replace_malloc.c:785)
==26535== by 0x54BA28E: virReallocN (viralloc.c:245)
==26535== by 0x54BA384: virExpandN (viralloc.c:294)
==26535== by 0x54BA657: virInsertElementsN (viralloc.c:436)
==26535== by 0x55DB011: virNWFilterObjListAssignDef (virnwfilterobj.c:362)
==26535== by 0x55DB530: virNWFilterObjListLoadConfig (virnwfilterobj.c:503)
==26535== by 0x55DB635: virNWFilterObjListLoadAllConfigs (virnwfilterobj.c:539)
==26535== by 0x2AC5A28B: nwfilterStateInitialize (nwfilter_driver.c:250)
==26535== by 0x5621C64: virStateInitialize (libvirt.c:770)
==26535== by 0x124379: daemonRunStateInit (libvirtd.c:881)
==26535== by 0x554AC78: virThreadHelper (virthread.c:206)
==26535== by 0x8F5F493: start_thread (in /lib64/libpthread-2.23.so)
Signed-off-by: Michal Privoznik <mprivozn@redhat.com>
Reviewed-by: John Ferlan <jferlan@redhat.com>
After bdcf6e481 there is a crasher in libvirt. The commit assumes
that priv->perf is always set. That is not true. For inactive
domains, the priv->perf is not allocated as it is set in
qemuProcessLaunch(). Now, usually we differentiate between
accesses to inactive and active definition and it works just
fine. Except for 'domstats'. There priv->perf is accessed without
prior check for domain inactivity. While we could check for that,
more robust solution is to make virPerfEventIsEnabled() accept
NULL.
How to reproduce:
1) ensure you have at least one inactive domain
2) virsh domstats
Signed-off-by: Michal Privoznik <mprivozn@redhat.com>
Reviewed-by: Daniel P. Berrange <berrange@redhat.com>
Even though there are several checks before calling this function
and for some scenarios we don't call it at all (e.g. on disk hot
unplug), it may be possible to sneak in some weird files (e.g. if
domain would have RNG with /dev/shm/some_file as its backend). No
matter how improbable, we shouldn't unlink it as we would be
unlinking a file from the host which we haven't created in the
first place.
Signed-off-by: Michal Privoznik <mprivozn@redhat.com>
Reviewed-by: Cedric Bosdonnat <cbosdonnat@suse.com>
Just like in previous commit, this fixes the same issue for
hotplug.
Signed-off-by: Michal Privoznik <mprivozn@redhat.com>
Reviewed-by: Cedric Bosdonnat <cbosdonnat@suse.com>
While the code allows devices to already be there (by some
miracle), we shouldn't try to create devices that don't belong to
us. For instance, we shouldn't try to create /dev/shm/file
because /dev/shm is a mount point that is preserved. Therefore if
a file is created there from an outside (e.g. by mgmt application
or some other daemon running on the system like vhostmd), it
exists in the qemu namespace too as the mount point is the same.
It's only /dev and /dev only that is different. The same
reasoning applies to all other preserved mount points.
Signed-off-by: Michal Privoznik <mprivozn@redhat.com>
Reviewed-by: Cedric Bosdonnat <cbosdonnat@suse.com>
Currently, all we need to do in qemuDomainCreateDeviceRecursive() is to
take given @device, get all kinds of info on it (major & minor numbers,
owner, seclabels) and create its copy at a temporary location @path
(usually /var/run/libvirt/qemu/$domName.dev), if @device live under
/dev. This is, however, very loose condition, as it also means
/dev/shm/* is created too. Therefor, we will need to pass more arguments
into the function for better decision making (e.g. list of mount points
under /dev). Instead of adding more arguments to all the functions (not
easily reachable because some functions are callback with strictly
defined type), lets just turn this one 'const char *' into a 'struct *'.
New "arguments" can be then added at no cost.
Signed-off-by: Michal Privoznik <mprivozn@redhat.com>
Reviewed-by: Cedric Bosdonnat <cbosdonnat@suse.com>
When setting up mount namespace for a qemu domain the following
steps are executed:
1) get list of mountpoints under /dev/
2) move them to /var/run/libvirt/qemu/$domName.ext
3) start constructing new device tree under /var/run/libvirt/qemu/$domName.dev
4) move the mountpoint of the new device tree to /dev
5) restore original mountpoints from step 2)
Note the problem with this approach is that if some device in step
3) requires access to a mountpoint from step 2) it will fail as
the mountpoint is not there anymore. For instance consider the
following domain disk configuration:
<disk type='file' device='disk'>
<driver name='qemu' type='raw'/>
<source file='/dev/shm/vhostmd0'/>
<target dev='vdb' bus='virtio'/>
<address type='pci' domain='0x0000' bus='0x00' slot='0x0a' function='0x0'/>
</disk>
In this case operation fails as we are unable to create vhostmd0
in the new device tree because after step 2) there is no /dev/shm
anymore. Leave aside fact that we shouldn't try to create devices
living in other mountpoints. That's a separate bug that will be
addressed later.
Currently, the order described above is rearranged to:
1) get list of mountpoints under /dev/
2) start constructing new device tree under /var/run/libvirt/qemu/$domName.dev
3) move them to /var/run/libvirt/qemu/$domName.ext
4) move the mountpoint of the new device tree to /dev
5) restore original mountpoints from step 3)
Signed-off-by: Michal Privoznik <mprivozn@redhat.com>
Reviewed-by: Cedric Bosdonnat <cbosdonnat@suse.com>
After 1eb6647979 nobody calls the iohelper with 6 arguments.
Everybody uses the other mode. Well, the only user of iohelper
after the previous commit is virFileWrapperFd really.
Signed-off-by: Michal Privoznik <mprivozn@redhat.com>
Reviewed-by: John Ferlan <jferlan@redhat.com>
Currently we use iohelper for virFDStream implementation. This is
because UNIX I/O can lie sometimes: even though a FD for a
file/block device is set as unblocking, actual read()/write() can
block. To avoid this, a pipe is created and one end is kept for
read/write while the other is handed over to iohelper to
write/read the data for us. Thus it's iohelper which gets blocked
and not our event loop.
This approach has two problems:
1) we are spawning a new process.
2) any exchange of information between daemon and iohelper can be
done only through the pipe.
Therefore, iohelper is replaced with an implementation in thread
which is created just for the stream lifetime. The data are still
transferred through pipe (for now), but both problems described
above are solved.
Signed-off-by: Michal Privoznik <mprivozn@redhat.com>
Reviewed-by: John Ferlan <jferlan@redhat.com>
While this is no functional change, it makes the code look a bit
nicer. Moreover, it prepares ground for future work.
Signed-off-by: Michal Privoznik <mprivozn@redhat.com>
Reviewed-by: John Ferlan <jferlan@redhat.com>
Michal Privoznik