libxl_set_memory_target seems to have the following return values:
'1' : on failure, if the failure happens because of a xenstore error
*or* invalid target
'-1': on error, the setmaxmem and set_pod_target hypercalls
return -1 and set errno appropriately.
'0' : on success
Make it consistently return ERROR_FAIL on failure, unless the
parameters were invalid, in which case return ERROR_INVAL.
In accordance with CODING_SYTLE:
1. Leave rc uninitialized, and set when an error is detected
2. Use 'r' for return values to functions whose return values are a
different error space (like xc_domain_setmaxmem and
xc_domain_set_pod_target)
3. Use 'lrc' for return values to local functions libxl__*
where a failure means retry, rather than fail the whole function
(libxl__fill_dom0_memory_info), to reduce the risk of that.
Signed-off-by: George Dunlap <George.Dunlap@eu.citrix.com>
Signed-off-by: Paulina Szubarczyk <paulinaszubarczyk@gmail.com>
Reviewed-by: Olaf Hering <olaf@aepfle.de>
Acked-by: Wei Liu <wei.liu2@citrix.com>
pygrub and qemuu need to be able to access a VM's disks locally in
order to be able to pull out the kernel and provide emulated disk
access, respectively. This can be done either by accessing the local
disk directly, or by plugging the target disk into dom0 to allow
access.
Unfortunately, while the plugging machinery works for pygrub, it does
not yet work for qemuu; this means that at the moment, disks with
hotplug scripts or disks with non-dom0 backends cannot be provided as
emulated devices to HVM domains.
Fortunately, disks using hotplug scripts created in dom0 do create a
block device as part of set-up, which can be accessed locally; and if
they use block-common.sh:write_dev, this path will be written to
physical-device-path.
Modify libxl__device_disk_setdefault() to be able to fish this path
out of xenstore and pass it to the caller.
Unfortunately, at the time pygrub runs, the devices have not yet been
set up. Rather than try to stash the domid somewhere to pass, we just
pass INVALID_DOMID.
This allows qemuu to emulate block devices created with custom hotplug
scripts.
Signed-off-by: George Dunlap <george.dunlap@eu.citrix.com>
Acked-by: Ian Jackson <ian.jackson@eu.citrix.com>
Tested-by: George Dunlap <george.dunlap@citrix.com>
qemu can also access disks which will be provided with a qdisk backend
directly; add a flag to libxl__device_disk_find_local_path to indicate
whether to check for qdisk direct access.
Call libxl__device_disk_find_local_path() for most paths. If we can't
find a local path, print an error and skip the disk, rather than using
a bogus path.
Now if there is no local access to the disk (i.e., because the disk
has a non-local backend, or relies on a custom hotplug script), libxl
will now print a warning and not provide the emulated disk, rather
than providing bogus parameters to qemu which cause it to error out.
(Such disks will still be available via the PV backend.)
I left the libxl__blktap_devpath in the qemuu-specific code rather
than sharing it with the pyrgub code because:
1) When the pygrub path runs the guest disks have not yet been set up
2) libxl__blktap_devpath() will give you the existing devpath if it
already exists, but will set one up for you if you don't. So on the
pygrub path, this would end up setting up a new tap device.
3) There is no tap-specific teardown code on the pygrub path, and I
don't want to add any (particularly since I'm hoping to remove tapdisk
altogether soon).
Signed-off-by: George Dunlap <george.dunlap@eu.citrix.com>
Acked-by: Ian Jackson <ian.jackson@eu.citrix.com>
Tested-by: George Dunlap <george.dunlap@citrix.com>
Move pygrub checks for local access ability into a separate function.
Also reorganize libxl__device_disk_local_initiate_attach so that we
don't initialize dls->disk unless we actually end up doing a local
attach.
Signed-off-by: George Dunlap <george.dunlap@eu.citrix.com>
Acked-by: Ian Jackson <ian.jackson@eu.citrix.com>
Tested-by: George Dunlap <george.dunlap@citrix.com>
Testing the hotplug external script path at the moment involves
actually setting up one of the alternate datapaths (blktap, iscsi,
&c). Simplify testing by making a script which does a simple loopback,
but still has a target that can't be used directly.
To use:
script=block-dummy,vdev=xvda,target=dummy:<file>
Signed-off-by: George Dunlap <george.dunlap@eu.citrix.com>
Acked-by: Ian Jackson <ian.jackson@eu.citrix.com>
Tested-by: George Dunlap <george.dunlap@citrix.com>
This makes xl more useful in scripts.
The strange thing about this is that the internal cd_insert function
*already* returned something appropriate, and cd-eject was using it,
but cd-insert wasn't.
Also:
* Rework cd_insert to return EXIT_FAILURE and EXIT_SUCCESS rather than
magic constants
* Use 'r' for non-libxl return code, as specified in CODING_STYLE
Signed-off-by: George Dunlap <george.dunlap@eu.citrix.com>
Acked-by: Ian Jackson <ian.jackson@eu.citrix.com>
Also move the rc -> shell code translation into set_memory_max() to
make the two functions consistent with each other, and with other
similar examples in xl_cmdimpl.c
Change a 'long long' to "int64_t" while we're at it.
Signed-off-by: George Dunlap <george.dunlap@eu.citrix.com>
Acked-by: Ian Jackson <ian.jackson@eu.citrix.com>
There's no obvious reason for the call to xc_domain_getinfolist -- all
it seems to be doing is checking that the domain exists; but if it
doesn't exist, it will have already failed by this point.
NB that this will change the return value for libxl_set_memory_target:
now it will return 0 on success, rather than returning 1 (which was
the previous behavior). This is more in line with expected behavior,
and also allows the caller to distingiush between success and other
failure modes (some of which also return 1).
Signed-off-by: George Dunlap <george.dunlap@citrix.com>
Acked-by: Ian Jackson <ian.jackson@eu.citrix.com>
find_vec() is supposed to find the vector in the list if it exists,
choose an empty slot if it doesn't exist, and return null if all slots
are full.
However, coverity noticed that although the callers of find_vec() handle
the last condition, find_vec() itself didn't.
Check to see if we actually found an empty slot before attempting to
initialize it.
CID 1306864
Signed-off-by: George Dunlap <george.dunlap@citrix.com>
Acked-by: Ian Jackson <ian.jackson@eu.citrix.com>
Coverity notices that if !head is true, that !N can never be true.
Don't bother checking N, since we know it has to be at least one.
CID 1354243
Signed-off-by: George Dunlap <george.dunlap@citrix.com>
Acked-by: Ian Jackson <ian.jackson@eu.citrix.com>
We have several outstanding patch series which add devices that have
two levels: a controller and individual devices attached to that
controller.
In the interest of consistency, this patch introduces a section that
sketches out a template for interfaces for such devices.
Signed-off-by: George Dunlap <george.dunlap@citrix.com>
Acked-by: Juergen Gross <jgross@suse.com>
Acked-by: Olaf Hering <olaf@aepfle.de>
Acked-by: Chun Yan Liu <cyliu@suse.com>
Acked-by: Ian Campbell <ian.campbell@citrix.com>
Changeset 1166ecf ('tools/Rules.mk: Don't optimize debug builds; add
macro debugging information') exposed a bug whereby the autoconf stuff
in tools was setting its own debug value (defaulting to ENABLED, even
for releases) instead of honoring the value set in Config.mk.
After that changeset, if the global build has -D_FORTIFY_SOURCE
enabled (as is the default in CentOS 7 rpmbuild), then the tools build
will fail (because debug builds default to on).
There should be only one place to specify whether to build debug or
not, and Config.mk is already included by the relevant makefiles. So
simply remove the tools/configure debug option and everything falls
into place naturally.
Signed-off-by: George Dunlap <george.dunlap@eu.citrix.com>
Acked-by: Ian Campbell <ian.campbell@citrix.com>
Acked-by: Wei Liu <wei.liu2@citrix.com>
The existing ept_set_entry() and ept_get_entry() routines are extended
to optionally set/get suppress_ve. Passing -1 will set suppress_ve on
new p2m entries, or retain suppress_ve flag on existing entries.
Signed-off-by: George Dunlap <george.dunlap@eu.citrix.com>
Signed-off-by: Ravi Sahita <ravi.sahita@intel.com>
Reviewed-by: Jan Beulich <jbeulich@suse.com>
Reviewed-by: George Dunlap <george.dunlap@eu.citrix.com>
Release-acked-by: Wei Liu <wei.liu2@citrix.com>
Also adjust the caller in set_identity_p2m_entry().
Signed-off-by: Jan Beulich <jbeulich@suse.com>
Try to avoid placing PCI BARs over RMRRs:
- If mmio_hole_size is not specified, and the existing MMIO range has
RMRRs in it, and there is space to expand the hole in lowmem without
moving more memory, then make the MMIO hole as large as possible.
- When placing RMRRs, find the next RMRR higher than the current base
in the lowmem mmio hole. If it overlaps, skip ahead of it and find
the next one.
This certainly won't work in all cases, but it should work in a
significant number of cases. Additionally, users should be able to
work around problems by setting mmio_hole_size larger in the guest
config.
Signed-off-by: George Dunlap <george.dunlap@eu.citrix.com>
Signed-off-by: Tiejun Chen <tiejun.chen@intel.com>
Reviewed-by: Jan Beulich <jbeulich@suse.com>
Release-acked-by: Wei Liu <wei.liu2@citrix.com>
The blocktap library isn't really necessary; all the necessary functionality
is available via the tap-ctl binary.
To use:
script=block-tap,vdev=[whatever],target=vhd:/path/to/file.vhd
script=block-tap,vdev=[whatever],target=aio:/path/to/file.raw
Signed-off-by: George Dunlap <george.dunlap@eu.citrix.com>
Acked-by: Wei Liu <wei.liu2@citrix.com>
They are no longer needed, having been replaced by a daemon for
driverdomains which will run scripts as necessary.
Worse yet, they seem to be broken for script-based block devices, such
as block-iscsi. This wouldn't matter so much if they were never run
by default; but if you run block-attach without having created a
domain, then the appropriate node to disable running udev scripts will
not have been written yet, and the attach will silently fail.
Rather than try to sort out that issue, just remove them entirely.
Signed-off-by: George Dunlap <george.dunlap@eu.citrix.com>
Acked-by: Wei Liu <wei.liu2@citrix.com>
There are a lot of paths through
libxl__device_disk_local_initiate_attach(), but they all really boil
down to one thing: Can we just access the file directly, or do we need
to attach it?
The requirements for direct access are fairly simple:
* Is this local (as opposed to a driver domain)?
* Is this a raw format (as opposed to cooked)?
* Does this have no scripts associated with it?
If it meets all those requirements, we can access it directly;
otherwise we need to attach it.
This fixes a bug where bootloader execution fails for disks with
hotplug scripts.
This should fix a theoretical bug when using a qdisk backend in a
driver domain. (Not tested.)
Based on a patch by Roger Pau Monne <roger.pau@citrix.com>.
Signed-off-by: George Dunlap <george.dunlap@eu.citrix.com>
Acked-by: Roger Pau Monné <roger.pau@citrix.com>
Acked-by: Ian Campbell <ian.campbell@citrix.com>
xentrace is the only caller at the moment. Split the cpu and event
mask setting out into seperate functions, but leave the current limit
of 32 bits for masks passed in from the command-line.
Based on a patch from Konrad Rzeszutek Wilk <konrad.wilk@oracle.com>
Acked-by: Ian Campbell <ian.campbell@citrix.com>
Signed-off-by: George Dunlap <george.dunlap@eu.citrix.com>
Signed-off-by: Konrad Rzeszutek Wilk <konrad.wilk@oracle.com>
Several users have reported that their available free memory in the
guest when they used maxmem >> memory was much smaller than when
maxmem == memory. This is the unavoidable consequence of how
ballooning works, but it's not something users expect.
Warn them of this effect in the place we tell them how to make it
happen, so they aren't surprised.
Signed-off-by: George Dunlap <george.dunlap@eu.citrix.com>
Acked-by: Ian Campbell <ian.campbell@citrix.com>
At the moment libxl unconditinally passes the underlying file format
to qemu in the device string. However, when tapdisk is in use,
tapdisk handles the underlying format and presents qemu with
effectively a raw disk. When qemu looks at the tapdisk block device
and doesn't find the image format it was looking for, it will fail.
This effectively means that tapdisk cannot be used with HVM domains at
the moment except for raw files.
Instead, if we're using a tapdisk backend, tell qemu to use a raw file
format.
Signed-off-by: George Dunlap <george.dunlap@eu.citrix.com>
Acked-by: Ian Campbell <ian.campbell@citrix.com>
Acked-by: Wei Liu <wei.liu2@citrix.com>
[ ijc -- nuked extra blank line ]
Return proper error codes on failure so that scripts can tell whether
the command completed properly or not.
This is not a proper fix, since it fails to call
libxl_device_disk_dispose() on the error path. But a proper fix
requires some refactoring, so given where we are in the release
process, it's better to have a fix that is simple and obvious, and do
the refactoring once the next development window opens up.
Signed-off-by: George Dunlap <george.dunlap@eu.citrix.com>
Acked-by: Ian Campbell <ian.campbell@citrix.com>
At the moment, making a release tarball is an annoyingly manual
process that involves running "git archive" into a temporary directory.
Script this process up and make a target, so that the release manager
can simply type "make src-tarball-release" and have everything show up
nice and neat in dist/xen-$version.tar.gz. "make src-tarball" will
make a version number based on git describe, which will typically have
the most recent tag, number of commits since that tag, and the git
commit id of the current HEAD.
Signed-off-by: George Dunlap <george.dunlap@eu.citrix.com>
subtree-force-update will update all subtrees according to the current TAG specified
in Config.mk.
Signed-off-by: George Dunlap <george.dunlap@eu.citrix.com>
Signed-off-by: Ian Jackson <ian.jackson@eu.citrix.com>
Make the targets depend on the "-find" targets, so that if the
subtrees are cloned if they haven't been cloned already.
Signed-off-by: George Dunlap <george.dunlap@eu.citrix.com>
Acked-by: Ian Jackson <ian.jackson@eu.citrix.com>
We have four different external trees, and four different naming
conventions for specifying the URL, where they live, and what revision
to check out.
Normalize config options on the following bases:
- QEMU_UPSTREAM
- QEMU_TRADITIONAL
- SEABIOS_UPSTREAM
- OVMF_UPSTREAM
The following suffixes for all trees:
- _URL : A remote repository to clone from
- _REVISION : The revision to check out
And the following suffixes for the qemu trees:
- _INTREE : The location of an inlined tree (for tarball releases)
- _LOC : Where to actually look (either a directory or a URL)
The following parameters are still supported for backwards
compatibility:
- CONFIG_QEMU (=> QEMU_TRADITIONAL_LOC)
- QEMU_REMOTE (=> QEMU_TRADITIONAL_URL)
- QEMU_TAG (=> QEMU_TRADITIONAL_REVISION)
- SEABIOS_UPSTREAM_TAG (=> SEABIOS_UPSTREAM_REVISION)
Signed-off-by: George Dunlap <george.dunlap@eu.citrix.com>
Acked-by: Ian Jackson <ian.jackson@eu.citrix.com>
Currently xen, qemu-xen, and qemu-xen-traditional are kept in separate
repositories, but when we release them as a tarball, qemu-xen and
qemu-xen-traditional are in-lined into the tools/ directory.
In order to make this "just work", at the moment developer doing the
release manually modifies Config.mk as part of the relase process so
that CONFIG_QEMU and QEMU_UPSTREAM_URL point into the tools/ directory
instead.
Modify Config.mk to automatically check there before trying a remote
repository.
Signed-off-by: George Dunlap <george.dunlap@eu.citrix.com>
Acked-by: Ian Jackson <ian.jackson@eu.citrix.com>
The man page mentions passing config snippets key=val in xl create.
Unfortunately, the config syntax contain characters which are
interpreted (and often discarded) by the shell before getting passed
in, resulting in a parsing error.
For example:
xl create hvm.cfg cpus="0-3"
The shell will eat the quotes and pass 'cpus=0-3' to xl, which won't
parse properly without the quotes.
Mention this in the man page, and recommend the use of single quotes, as well
as separating multiple arguments with a semicolon, thus:
xl create hvm.cfg 'cpus="0-3"; pci=["1.0","1.1"]'
Signed-off-by: George Dunlap <george.dunlap@eu.citrix.com>
Acked-by: Ian Jackson <ian.jackson@eu.citrix.com>
The "seize" option tells the toolstack to attempt to automatically
unbind devices and re-bind them to the pciback driver. This should
make creating VMs that habitually use pass-through (such as driver domain
VMs and gaming VMs) easier to use and manage.
[Whitespace error fixed by iwj.]
Signed-off-by: George Dunlap <george.dunlap@eu.citrix.com>
The 30ms timeslice was chosen nearly a decade ago now, with cpu
"burning" workloads in mind. In the mean time, processors have gotten
faster and VMEXITs have gotten faster. A timeslice of 30ms has a
major cost when running latency-sensitive workloads like network or
audio streaming: getting caught behind just one or two other VMs can
introduce a processing delay of up to 60ms, and the "round-robin"
nature of the credit scheduler means this delay may be introduced
every time the VM yields for periods of time.
The XenServer performance team at Citrix have done extensive testing
with various timeslices, including 30ms, 10ms, 5ms, and 2ms. None of
the workloads exhibited any performance degradation with a 5ms
timeslice.
Signed-off-by: George Dunlap <george.dunlap@eu.citrix.com>
Reviewed-by: Dario Faggioli <dario.faggioli@citrix.com>
Acked-by: Tim Deegan <tim@xen.org>
Register an IO handler for the entire PIO range, and have it call the
PV PIO handlers.
NB at this point this won't do the full "copy and execute on the stack
with full GPRs" work-around; this may need to be sorted out for dom0 to allow
these instructions to happen in guest context.
Signed-off-by: George Dunlap <george.dunlap@eu.citrix.com>
Reviewed-by: Jan Beulich <jbeulich@suse.com>
Acked-by: Keir Fraser <keir@xen.org>
Acked-by: Eddie Dong <eddie.dong@intel.com>
PVH guests don't have a backing device model emulator (qemu); just
tolerate this situation explicitly, rather than special-casing PVH.
For unhandled IO, hvmemul_do_io() will now return X86EMUL_OKAY, which
is I believe what would be the effect if qemu didn't have a handler
for the IO.
This also fixes a potetial DoS in the host from the reworked series:
If the guest makes a hypercall which sends an invalidate request, it
would have crashed the host.
Signed-off-by: George Dunlap <george.dunlap@eu.citrix.com>
Reviewed-by: Jan Beulich <jbeulich@suse.com>
Acked-by: Keir Fraser <keir@xen.org>
Acked-by: Eddie Dong <eddie.dong@intel.com>
Removing the assert allows the PVH code to call this during vmcs
construction in a later patch, making the code more robust by removing
duplicate code.
Signed-off-by: George Dunlap <george.dunlap@eu.citrix.com>
Acked-by: Tim Deegan <tim@xen.org>
Reviewed-by: Jan Beulich <jbeulich@suse.com>
Acked-by: Keir Fraser <keir@xen.org>
Acked-by: Eddie Dong <eddie.dong@intel.com>
If the allocation fails, make sure to call vmx_vmcs_exit().
This is a candidate for backport.
Signed-off-by: George Dunlap <george.dunlap@eu.citrix.com>
Signed-off-by: Mukesh Rathor <mukesh.rathor@oracle.com>
It looks like one of the failure cases in hvm_vcpu_initialise jumps to
the wrong label; this could lead to slow leaks if something isn't
cleaned up properly.
I will probably change these labels in a future patch, but I figured
it was better to have this fix separately.
This is also a candidate for backport.
Signed-off-by: George Dunlap <george.dunlap@eu.citrix.com>
Signed-off-by: Mukesh Rathor <mukesh.rathor@oracle.com>
xm block-attach takes "r" as read-only, not "ro".
This is a backport candidate for 4.3, and probably others.
Signed-off-by: George Dunlap <george.dunlap@eu.citrix.com>
Acked-by: Ian Campbell <ian.campbell@citrix.com>
At the moment, qemu-xen can't handle memory being relocated by
hvmloader. This may happen if a device with a large enough memory
region is passed through to the guest. At the moment, if this
happens, then at some point in the future qemu will crash and the
domain will hang. (qemu-traditional is fine.)
It's too late in the release to do a proper fix, so we try to do
damage control.
hvmloader already has mechanisms to relocate memory to 64-bit space if
it can't make a big enough MMIO hole. By default this is 2GiB; if we
just refuse to make the hole bigger if it will overlap with guest
memory, then the relocation will happen by default.
v5:
- Update comment to not refer to "this series".
v4:
- Wrap long line in libxl_dm.c
- Fix comment
v3:
- Fix polarity of comparison
- Move diagnostic messages to another patch
- Tested with xen platform pci device hacked to have different BAR sizes
{256MiB, 1GiB} x {qemu-xen, qemu-traditional} x various memory
configurations
- Add comment explaining why we default to "allow"
- Remove cast to bool
v2:
- style fixes
- fix and expand comment on the MMIO hole loop
- use "%d" rather than "%s" -> (...)?"1":"0"
- use bool instead of uint8_t
- Move 64-bit bar relocate detection to another patch
- Add more diagnostic messages
Signed-off-by: George Dunlap <george.dunlap@eu.citrix.com>
Acked-by: Ian Jackson <ian.jackson@eu.citrix.com>
CC: Ian Campbell <ian.campbell@citrix.com>
CC: Stefano Stabellini <stefano.stabellini@citrix.com>
CC: Hanweidong <hanweidong@huawei.com>
CC: Keir Fraser <keir@xen.org>
CC: Keir Fraser <keir@xen.org>
Allow devices with BARs less than 512MiB to be relocated to high
memory.
This will only be invoked if there is not enough low MMIO space to map
the device, and will be done preferentially to large devices first; so
in all likelihood only large devices will be remapped anyway.
This is needed to work-around the issue of qemu-xen not being able to
handle moving guest memory around to resize the MMIO hole. The
default MMIO hole size is less than 256MiB.
v3:
- Fixed minor style issue
Signed-off-by: George Dunlap <george.dunlap@eu.citrix.com>
Reviewed-by: Jan Beulich <jbeulich@suse.com>
Acked-by: Stefano Stabellini <stefano.stabellini@eu.citrix.com>
Acked-by: Ian Jackson <ian.jackson@eu.citrix.com>
CC: Ian Campbell <ian.campbell@citrix.com>
CC: Hanweidong <hanweidong@huawei.com>
CC: Keir Fraser <keir@xen.org>
Keep track of how much mmio space is left total, as well as the amount
of "low" MMIO space (<4GiB), and only load devices into high memory if
there is not enough low memory for the rest of the devices to fit.
Because devices are processed by size in order from large to small,
this should preferentially relocate devices with large BARs to 64-bit
space.
v3:
- Just use mmio_total rather than introducing a new variable.
- Port to using mem_resource directly rather than low_mmio_left
Signed-off-by: George Dunlap <george.dunlap@eu.citrix.com>
Reviewed-by: Jan Beulich <jbeulich@suse.com>
Acked-by: Stefano Stabellini <stefano.stabellini@eu.citrix.com>
Acked-by: Ian Jackson <ian.jackson@eu.citrix.com>
CC: Ian Campbell <ian.campbell@citrix.com>
CC: Stefano Stabellini <stefano.stabellini@citrix.com>
CC: Hanweidong <hanweidong@huawei.com>
CC: Keir Fraser <keir@xen.org>
When deciding whether to map a device in low MMIO space (<4GiB),
hvmloader compares it with "mmio_left", which is set to the size of
the low MMIO range (pci_mem_end - pci_mem_start). However, even if it
does map a device in high MMIO space, it still removes the size of its
BAR from mmio_left.
In reality we don't need to do a separate accounting of the low memory
available -- this can be calculated from mem_resource. Just get rid
of the variable and the duplicate accounting entirely. This will make
the code more robust.
Note also that the calculation of whether to move a device to 64-bit
is fragile at the moment, depending on some unstated assumptions.
State those assumptions in a comment for future reference.
v5:
- Add comment documenting fragility of the move-to-highmem check
v3:
- Use mem_resource values directly instead of doing duplicate
accounting
Signed-off-by: George Dunlap <george.dunlap@eu.citrix.com>
Reviewed-by: Jan Beulich <jbeulich@suse.com>
Acked-by: Stefano Stabellini <stefano.stabellini@eu.citrix.com>
Acked-by: Ian Jackson <ian.jackson@eu.citrix.com>
CC: Ian Campbell <ian.campbell@citrix.com>
CC: Stefano Stabellini <stefano.stabellini@citrix.com>
CC: Hanweidong <hanweidong@huawei.com>
CC: Keir Fraser <keir@xen.org>
After attempting to resize the MMIO hole, the check to determine
whether there is a need to relocate BARs into 64-bit space checks the
specific thing that caused the loop to exit (MMIO hole == 2GiB) rather
than checking whether the required MMIO will fit in the hole.
But even then it does it wrong: the polarity of the check is
backwards.
Check for the actual condition we care about (the sizeof the MMIO
hole) rather than checking for the loop exit condition.
v3:
- Move earlier in the series, before other functional changes
Signed-off-by: George Dunlap <george.dunlap@eu.citrix.com>
Reviewed-by: Jan Beulich <jbeulich@suse.com>
Acked-by: Stefano Stabellini <stefano.stabellini@eu.citrix.com>
Acked-by: Ian Jackson <ian.jackson@eu.citrix.com>
CC: Ian Jackson <ian.jackson@citrix.com>
CC: Ian Campbell <ian.campbell@citrix.com>
CC: Stefano Stabellini <stefano.stabellini@citrix.com>
CC: Hanweidong <hanweidong@huawei.com>
CC: Keir Fraser <keir@xen.org>
hvmloader will read hvm_info->high_mem_pgend to calculate where to
start the highmem PCI region. However, if the guest does not have any
memory in the high region, this is set to zero, which will cause
hvmloader to use the "0" for the base of the highmem region, rather
than 1 << 32.
Check to see whether hvm_info->high_mem_pgend is set; if so, do the
normal calculation; otherwise, use 1<<32.
v4:
- Handle case where hfm_info->high_mem_pgend is non-zero but doesn't
point into high memory, throwing a warning.
Signed-off-by: George Dunlap <george.dunlap@eu.citrix.com>
Acked-by: Ian Jackson <ian.jackson@eu.citrix.com>
Acked-by: Stefano Stabellini <stefano.stabellini@eu.citrix.com>
CC: Ian Campbell <ian.campbell@citrix.com>
CC: Ian Jackson <ian.jackson@citrix.com>
CC: Stefano Stabellini <stefano.stabellini@citrix.com>
CC: Hanweidong <hanweidong@huawei.com>
CC: Keir Fraser <keir@xen.org>
The qxl drivers for Windows and Linux end up calling instructions
that cannot be used for MMIO at the moment. Just for the 4.3 release,
remove qxl support.
This patch should be reverted as soon as the 4.4 development window opens.
The issue in question:
(XEN) emulate.c:88:d18 bad mmio size 16
(XEN) io.c:201:d18 MMIO emulation failed @ 0033:7fd2de390430: f3 0f 6f
19 41 83 e8 403
The instruction in question is "movdqu (%rcx),%xmm3". Xen knows how
to emulate it, but unfortunately %xmm3 is 16 bytes long, and the interface
between Xen and qemu at the moment would appear to only allow MMIO accesses
of 8 bytes.
It's too late in the release cycle to find a fix or a workaround.
Signed-off-by: George Dunlap <george.dunlap@eu.citrix.com>
Acked-by: Ian Jackson <ian.jackson@eu.citrix.com>
Acked-by: Ian Campbell <ian.campbell@citrix.com>
We've changed the config option from "vifscript" to "vif.default.script". This
was changed in the manpage but not in the example xl.conf.
Also move the option down to be with the other vif options.
Signed-off-by: George Dunlap <george.dunlap@eu.citrix.com>
Acked-by: Ian Campbell <ian.campbell@citrix.com>
Linux uses sse4_1 and sse4_2, but at the moment libxl uses '.' instead
of '_'. This makes it confusing for people looking in Linux's /proc/cpuinfo
to disable features.
Add the Linux feature names, keeping the old ones for compatability.
Signed-off-by: George Dunlap <george.dunlap@eu.citrix.com>
Acked-by: Ian Campbell <ian.camppbell@citrix.com>
Two changes:
* Stat the file before calling libxl_cdrom_insert()
* Return an error if anything fails (including libxl_cdrom_insert)
This is in part to work around the fact that the RAW disk type
is used for things that aren't actually files; so we can't call
stat in libxl_device.c:libxl__device_disk_set_backend() because
it may be going over a remote protocol.
Signed-off-by: George Dunlap <george.dunlap@eu.citrix.com>
Acked-by: Ian Campbell <ian.campbell@citrix.com>
Allow the "usbdevice" key to accept a list of USB devices, and pass
them in using the new usbdevice_list domain build element.
For backwards compatibility, still accept singleton values.
Also update the xl.cfg manpage, adding information about how to pass
through host devices.
as applied:
- Fix trailing whitespace and wrap some lines in xl_cmdimpl.c -iwj
v2:
- Add some verbiage to make it clear that "usb" is for emulated devices
- Reference qemu manual for more usbdevice options
Signed-off-by: George Dunlap <george.dunlap@eu.citrix.com>
Acked-by: Ian Jackson <ian.jackson@eu.citrix.com>
Signed-off-by: Ian Jackson <ian.jackson@eu.citrix.com>
This patch allows an HVM domain to be created with multiple USB
devices.
Since the previous interface only allowed the passing of a single
device, this requires us to add a new element to the hvm struct of
libxl_domain_build_info -- usbdevice_list. For API compatibility, the
old element, usbdevice, remains.
If hvm.usbdevice_list is set, each device listed will cause an extra
"-usbdevice [foo]" to be appended to the qemu command line.
Callers may set either hvm.usbdevice or hvm.usbdevice_list, but not
both; libxl will throw an error if both are set.
In order to allow users of libxl to write software compatible with
older versions of libxl, also define LIBXL_HAVE_BUILDINFO_USBDEVICE_LIST.
If this is defined, callers may use either hvm.usbdevice or
hvm.usbdevice_list; otherwise, only hvm.usbdevice will be available.
as applied:
- Fix whitespace errors -iwj
v3:
- Duplicate functionality in both "new" and "old", since we're not
unifying the two anymore.
v2:
- Throw an error if both usbdevice and usbdevice_list are set
- Update and clarify definition based on feedback
- Previous patches means this works for both traditional and upstream
Signed-off-by: George Dunlap <george.dunlap@eu.citrix.com>
Acked-by: Ian Jackson <ian.jackson@eu.citrix.com>
Signed-off-by: Ian Jackson <ian.jackson@eu.citrix.com>
Makes the following changes to the vnc generation code:
* Simplifies and comments it, making it easier to read and grok
* Throws an error if duplicate values of display are set, rather
than the current very un-intuitive behavior.
Signed-off-by: George Dunlap <george.dunlap@eu.citrix.com>
Acked-by: Ian Jackson <ian.jackson@eu.citrix.com>
Under normal circumstances, snext->credit should never be less than
-CSCHED_MIN_TIMER. However, under some circumstances, a vcpu with low
credits may be allowed to run long enough that its credits are
actually less than -CSCHED_CREDIT_INIT.
(Instances have been observed, for example, where a vcpu with 200us of
credit was allowed to run for 11ms, giving it -10.8ms of credit. Thus
it was still negative even after the reset.)
If this is the case for snext, we simply want to keep moving everyone
up until it is in the black again. This fair because none of the
other vcpus want to run at the moment.
Rather than loop, just detect how many times we want to add
CSCHED_CREDIT_INIT. Try to avoid integer divides and multiplies in
the common case.
Signed-off-by: George Dunlap <george.dunlap@eu.citrix.com>
In order to avoid high-frequency cpu migration, vcpus may in fact be
scheduled slightly out-of-order. Account for this situation properly.
Signed-off-by: George Dunlap <george.dunlap@eu.citrix.com>
It's probably a good idea to re-evaluate placement whenever the
affinity changes.
This additionally has the benefit of removing scheduler-specific
exceptions introduced in git c/s e6a6fd63.
The conditionals surrounding vcpu_migrate() are left pending a re-work
of the logic to avoid the common case calling vcpu_migrate() twice (once
here, and once in context_saved().
Signed-off-by: George Dunlap <george.dunlap@eu.citrix.com>
This should help with under-accounting of vCPU-s running for extremly
short periods of time, but becoming runnable again at a high frequency.
Don't bother subtracting the residual from the runtime, as it can only ever
add up to one nanosecond, and will end up being debited during the next
reset interval anyway.
Original-patch-by: Jan Beulich <jbeulich@suse.com>
Signed-off-by: George Dunlap <george.dunlap@eu.citrix.com>
csched_runtime() needs to call the ct2() function to change credits
into time. The c2t() function, however, is expensive, as it requires
an integer division.
c2t() was being called twice, once for the main vcpu's credit and once
for the difference between its credit and the next in the queue. But
this is unnecessary; by calculating in "credit" first, we can make it
so that we just do one conversion later in the algorithm.
This also adds more documentation describing the intended algorithm,
along with a relevant assertion..
The effect of the new code should be the same as the old code.
Spotted-by: Jan Beulich <JBeulich@suse.com>
Signed-off-by: George Dunlap <george.dunlap@eu.citrix.com>
The flags structure is not protected by locks (or more precisely,
it is protected using an inconsistent set of locks); we therefore need
to make sure that all accesses are atomic-safe. This is particulary
important in the case of the PARKED flag, which if clobbered while
changing the YIELD bit will leave a vcpu wedged in an offline state.
Using the atomic bitops also requires us to change the size of the "flags"
element.
Spotted-by: Igor Pavlikevich <ipavlikevich@gmail.com>
Signed-off-by: George Dunlap <george.dunlap@eu.citrix.com>
libxl__device_disk_from_xs_be() was failing without error for some
missing xenstore nodes in a backend, while assuming (without checking)
that other nodes were valid, causing a crash when another internal
error wrote these nodes in the wrong place.
Make this function consistent by:
* Checking the existence of all nodes before using
* Choosing a default only when the node is not written in device_disk_add()
* Failing with log msg if any node written by device_disk_add() is not present
* Returning an error on failure
* Disposing of the structure before returning using libxl_device_disk_displose()
Also make the callers of the function pay attention to the error and
behave appropriately. In the case of libxl__append_disk_list_of_type(),
this means only incrementing *ndisks as the disk structures are
successfully initialized.
Signed-off-by: George Dunlap <george.dunlap@eu.citrix.com>
Acked-by: Ian Campbell <ian.campbell@citrix.com>
Committed-by: Ian Campbell <ian.campbell@citrix.com>
If the user has set a vnc display number both in vnclisten (with
"xxxx:yy"), and with vncdisplay, throw an error.
Update man pages to match.
Signed-off-by: George Dunlap <george.dunlap@eu.citrix.com>
Acked-by: Ian Campbell <ian.campbell@citrix.com>
Committed-by: Ian Campbell <ian.campbell@citrix.com>
libxl_cdrom_insert was failing to initialize the backend type,
resulting in the wrong default backend. The result was not only that
the CD was not inserted properly, but also that some improper xenstore
entries were created, causing further block commands to fail.
This patch fixes the bug by setting the disk backend type based on the
type of the existing device.
It also makes the system more robust by checking to see that it has
got a valid path before proceeding to write a partial xenstore entry.
Signed-off-by: George Dunlap <george.dunlap@eu.citrix.com>
Acked-by: Ian Campbell <ian.campbell@citrix.com>
Committed-by: Ian Campbell <ian.campbell@citrix.com>
libxl_cpupool_list() enumerates the cpupools by "probing": calling
cpupool_info, starting at 0 and stopping when it gets an error. However,
cpupool_info will print an error when the call to xc_cpupool_getinfo() fails,
resulting in every xl command that uses libxl_list_cpupool (such as
cpupool-list) printing that error message spuriously.
Since at the moment the times we want to print the message correspond
with the use of the existing "exact" parameter, use it to decide
whether to print the message or not.
Signed-off-by: George Dunlap <george.dunlap@eu.citrix.com>
Acked-by: Ian Campbell <ian.campbell@citrix.com>
Committed-by: Ian Campbell <ian.campbell@citrix.com>
The sched_credit_default_yield option was added when the behavior of
"SCHEDOP_yield" was changed in 4.1, to allow any users who had
problems to revert to the old behavior. The new behavior has been in
Xen.org xen since 4.1, and in XenServer even longer, and there is no
evidence of anyone having trouble with it. Remove the option.
Signed-off-by: George Dunlap <george.dunlap@eu.citrix.com>
Committed-by: Keir Fraser <keir@xen.org>
* Give reason for check (unpredictable results)
* Give a better recommendation (shut down xend)
* Make it clear that -f is overriding a safety check.
Signed-off-by: George Dunlap <george.dunlap@eu.citrix.com>
Acked-by: Ian Jackson <ian.jackson@eu.citrix.com>
Committed-by: Ian Jackson <ian.jackson@eu.citrix.com>
Testing has shown that doing sweeps for superpages slows down boot
significantly, but does not result in a significantly higher number of
superpages after boot. Early sweeping for 4k pages causes superpages
to be broken up unnecessarily.
Only sweep if we're really out of memory.
v2:
- Move unrelated code-motion hunk to another patch
v3:
- Remove now-unused reclaim_super from pod struct
Signed-off-by: George Dunlap <george.dunlap@eu.citrix.com>
Acked-by: Tim Deegan <tim@xen.org>
Committed-by: Tim Deegan <tim@xen.org>
When demand-populating pages due to guest accesses, check recently populated
pages to see if we can reclaim them for the cache. This should keep the PoD
cache filled when the start-of-day scrubber is going through.
The number 128 was chosen by experiment. Windows does its page
scrubbing in parallel; while a small nubmer like 4 works well for
single VMs, it breaks down as multiple vcpus are scrubbing different
pages in parallel. Increasing to 128 works well for higher numbers of
vcpus.
v2:
- Wrapped some long lines
- unsigned int for index, unsigned long for array
v3:
- Use PAGE_ORDER_2M instead of 9
- Removed inappropriate use of p2m_pod_zero_check_superpage() return value
Signed-off-by: George Dunlap <george.dunlap@eu.citrix.com>
Acked-by: Tim Deegan <tim@xen.org>
Committed-by: Tim Deegan <tim@xen.org>
Windows balloon drivers can typically only get 4k pages from the kernel,
and so hand them back at that level. Try to regain superpages by checking
the superpage frame that the 4k page is in to see if we can reclaim the whole
thing for the PoD cache.
This also modifies p2m_pod_zero_check_superpage() to return SUPERPAGE_PAGES on
success.
v2:
- Rewritten to simply to the check as in demand-fault case, without needing
to know that the p2m entry is a superpage.
- Also, took out the re-writing of the reclaim loop, leaving it optimized for
4k pages (by far the most common case), and simplifying the patch.
v3:
- Add SoB
Signed-off-by: George Dunlap <george.dunlap@eu.citrix.com>
Acked-by: Tim Deegan <tim@xen.org>
Committed-by: Tim Deegan <tim@xen.org>
Because of conflicts with another patch, changeset 25344:0f3b1e13d6af missed
one change.
Signed-off-by: George Dunlap <george.dunlap@eu.citrix.com>
Committed-by: Ian Campbell <ian.campbell@citrix.com>
pci-assignable-add will always store the driver rebind path, but
pci-assignable-remove will only actually rebind if asked to do so.
Signed-off-by: George Dunlap <george.dunlap@eu.citrix.com>
Committed-by: Ian Campbell <ian.campbell@citrix.com>
Introduce libxl helper functions to prepare devices to be passed
through to guests. This is meant to replace of all the manual sysfs
commands which are currently required.
pci_assignable_add accepts a BDF for a device and will:
* Unbind a device from its current driver, if any
* If "rebind" is set, it will store the path of the driver from which we
unplugged it in /libxl/pciback/$BDF/driver_path
* If create a slot for it in pciback if one doesn't yet exist
* Bind the device to pciback
At this point it will show up in pci_assignable_list, and is ready to
be passed through to a guest.
pci_assignable_remove accepts a BDF for a device and will:
* Unbind the device from pciback
* Remove the slot from pciback
* If "rebind" is set, and /libx/pciback/$BDF/driver_path exists, it
will attempt to rebind the device to its original driver.
Both functions are idempotent: if the desired end state has already
been reached, they return SUCCESS.
NB that "$BDF" in this case uses '-' instead of ':' and '.', because
':' and '.' are illegal characters in xenstore paths.
Signed-off-by: George Dunlap <george.dunlap@eu.citrix.com>
Committed-by: Ian Campbell <ian.campbell@citrix.com>
...to prepare for a consistent "pci_assignable_*" naming scheme.
Also move the man page entry into the PCI PASS-THROUGH section, rather
than the XEN HOST section.
No functional changes.
Signed-off-by: George Dunlap <george.dunlap@eu.citrix.com>
Acked-by: Ian Campbell <ian.campbell@citrix.com>
Committed-by: Ian Campbell <ian.campbell@citrix.com>
Many places in xl there's a variable or element named "filename" which
does not contain a filename, but the source of the data for reporting
purposes. Worse, there are variables which are sometimes actually
used or interpreted as a filename depending on what other variables
are set. This makes it difficult to tell when a string is purely
cosmetic, and when another bit of code may actually attempt to call
"open" with the string.
This patch makes a consistent distinction between "filename" (which
always refers to the name of an actual file, and may be interpreted as
such at some point) and "source" (which may be a filename, or may be
another data source such as a migration stream or saved data).
This does add some variables and reshuffle where assignments happen;
most notably, the "restore_filename" element of struct domain_create
is now only set when restoring from a file.
But at a high level, there should be no funcitonal changes.
Signed-off-by: George Dunlap <george.dunlap@eu.citrix.com>
Committed-by: Ian Campbell <ian.campbell@citrix.com>
The "filename" is a bit of a misnomer, as it's only used during error
messages, and in most instances cases is actually set to "command
line".
Rename it to "config_source" to make it clear that it's not used to
actually open any files.
No functional changes.
Signed-off-by: George Dunlap <george.dunlap@eu.citrix.com>
Committed-by: Ian Campbell <ian.campbell@citrix.com>
pygrub and xenpvnetboot are meant to be run by tools, and not by the user,
and thus should be in /usr/lib/xen/bin rather than /usr/bin.
Because most config files will still have an absolute path pointing to
/usr/bin/pygrub, make a symbolic link that we will deprecate.
Signed-off-by: George Dunlap <george.dunlap@eu.ctirix.com>
Acked-by: Ian Campbell <ian.campbell@citrix.com>
Committed-by: Ian Campbell <ian.campbell@citrix.com>
If the full path for a bootloader (such as pygrub or xenpvnetboot) is not
given, check for it first in the libexec path before falling back to the
PATH variable.
Signed-off-by: George Dunlap <george.dunlap@eu.citrix.com>
Committed-by: Ian Campbell <ian.campbell@citrix.com>
This involves making a goto clean-up path, rather than calling
return directly.
Signed-off-by: George Dunlap <george.dunlap@eu.citrix.com>
Acked-by: Ian Jackson <ian.jackson@eu.citrix.com>
Committed-by: Ian Jackson <ian.jackson.citrix.com>
Since the key information can be fairly simply put on the command-line,
there's no need to require an actual config file.
Also improve the help to cross-reference the xlcpupool.cfg manpage.
Signed-off-by: George Dunlap <george.dunlap@eu.citrix.com>
Committed-by: Ian Jackson <ian.jackson.citrix.com>
Acked-by: Ian Jackson <ian.jackson@eu.citrix.com>
Currently, if "xl cpupool-create" is called with no cpus configured,
xl will choose a cpu at random from the list of unassigned cpus, and
if no unassigned cpus are available, it will fail.
This seems to me to be a poor interface. For one, it makes it impossible
to create an empty cpupool using the xl command-line, except by creating
a pool and then removing the cpus from it. For two, I don't think assigning
a random cpu is a feature; it's not unreasonable for the user to specify
which cpus to add to which pools.
This patch changes the behavior of "xl cpupool-create" to create an empty
pool if no cpus are specified. I believe this interface to be more expected
and more script-friendly.
Signed-off-by: George Dunlap <george.dunlap@eu.citrix.com>
Committed-by: Ian Jackson <ian.jackson.citrix.com>
Acked-by: Ian Jackson <ian.jackson@eu.citrix.com>
George Dunlap