Constantly m(un)locking memory can have significant overhead on
systems with large numbers of CPUs. This was previously fixed by
20841:fbe8f32fa257 but this was dropped during the transition to
hypercall buffers.
Introduce a small cache of single page hypercall buffer allocations
which can be resused to avoid this overhead.
Add some statistics tracking to the hypercall buffer allocations.
The cache size of 4 was chosen based on these statistics since they
indicated that 2 pages was sufficient to satisfy all concurrent single
page hypercall buffer allocations seen during "xl create", "xl
shutdown" and "xl destroy" of both a PV and HVM guest therefore 4
pages should cover the majority of important cases.
This fixes http://bugzilla.xensource.com/bugzilla/show_bug.cgi?id=1719.
Signed-off-by: Ian Campbell <ian.campbell@citrix.com>
Reported-by: Zheng, Shaohui <shaohui.zheng@intel.com>
Tested-by: Haitao Shan <maillists.shan@gmail.com>
Committed-by: Ian Jackson <ian.jackson@eu.citrix.com>
It's not clear why a userspace lzma decode would want to use that
particular value, what bearing it has on anything or why it would
assume it could use 1/3 of the total RAM in the system (potentially
quite a large amount of RAM) as opposed to any other limit number.
Instead, hardcode 32Mby.
This reverts 22830:c80960244942, removes the xc_get_physmem/physmem
function entirely, and replaces the expression at the call site with a
fixed constant.
Signed-off-by: Ian Jackson <ian.jackson@eu.citrix.com>
Acked-by: Ian Campbell <Ian.Campbell@eu.citrix.com>
Cc: Christoph Egger <Christoph.Egger@amd.com>
Committed-by: Ian Jackson <ian.jackson@eu.citrix.com>
NetBSD doesn't have sysconf(_SC_PHYS_PAGES).
Factor physmem() out into os-dependent files and rename it to
xc_get_physmem() so as not to pollute the namespace.
Signed-off-by: Christoph Egger <Christoph.Egger@amd.com>
Acked-by: Ian Jackson <ian.jackson@eu.citrix.com>
Committed-by: Ian Jackson <ian.jackson@eu.citrix.com>
Transition to xc_osdep_handle is now complete and nothing uses
(or should be using) it.
Signed-off-by: Ian Campbell <ian.campbell@citrix.com>
Signed-off-by: Ian Jackson <ian.jackson.citrix.com>
This patch introduces the basic infrastructure and uses it for open
and close operations on privcmd, evtchn and gnttab devices.
Signed-off-by: Ian Campbell <ian.campbell@citrix.com>
Signed-off-by: Ian Jackson <ian.jackson.citrix.com>
The xc_interface previously passed to xc_gnttab_* was only used for
logging which can now be done via the xc_gnttab handle instead.
This makes the interface consistent with the changes made to the main
interface in 21483:779c0ef9682c.
Also update QEMU_TAG to pull in the corresponding qemu change.
Signed-off-by: Ian Campbell <ian.campbell@citrix.com>
Signed-off-by: Ian Jackson <ian.jackson@eu.citrix.com>
This makes the interface consistent with the changes made to the main
interface in 21483:779c0ef9682c.
Also fix some references to "struct xc_interface" which should have
been simply "xc_interface" in tools/xenpaging, and update QEMU_TAG to
pull in the corresponding qemu change.
Signed-off-by: Ian Campbell <ian.campbell@citrix.com>
Signed-off-by: Ian Jackson <ian.jackson@eu.citrix.com>
(defer save/restore and shadow related interfaces til a later patch)
Signed-off-by: Ian Campbell <ian.campbell@citrix.com>
Signed-off-by: Ian Jackson <ian.jackson.citrix.com>
(Modified version of this patch, with updated #includes in
xc_hcall_buf.c to fix stubdomain build.)
Signed-off-by: Ian Campbell <ian.campbell@citrix.com>
Signed-off-by: Ian Jackson <ian.jackson@eu.citrix.com>
Now that all XENMEM_* callers go via an op specific function make
xc_memory_op private to libxc (and rename to do_memory_op for
consistency with other private functions).
Also change the interface to take a size parameter so that
do_memory_op knows how much memory to lock for the top-level argument,
removing some of the introspection.
[ Compatibility note: this breaks qemu-xen before 706bc8c588cb ]
Signed-off-by: Ian Campbell <ian.campbell@citrix.com>
Signed-off-by: Ian Jackson <ian.jackson@eu.citrix.com>
Not actually used here but useful to confirm that a handle is passed
down to each location where it will be required once we switch to
hypercall buffers.
Signed-off-by: Ian Campbell <ian.campbell@citrix.com>
Signed-off-by: Ian Jackson <ian.jackson@eu.citrix.com>
# User Ian Campbell <ian.campbell@citrix.com>
# Date 1282671421 -3600
# Node ID d1dd29a470ef1b9d2c77478a123326036dfe90bb
# Parent d7a4adad9c328decbd384d87b23001aea8951b86
tools/libxc, tools/libelf: Relicense under LGPL v2.1
Relicense these two libraries under LGPL v2.1 only except where
individual files already included the "or later" provision.
Copyright holders have been contacted by Stephen Spector and have all
agreed this change.
Removed tools/libxc/ia64/aclinux.h since it appeared to be
unused. There is a separate, more up to date, copy in
xen/include/acpi/platform/aclinux.h which does appear to be used.
Clarify the license of MiniOS privcmd.h under the same terms as other
tools/include/xen-sys headers.
Signed-off-by: Ian Campbell <ian.campbell@citrix.com>
Acked-by: Stephen Spector <stephen.spector@citrix.com>
Signed-off-by: Ian Jackson <ian.jackson@eu.citrix.com>
* Make "read_exact" in libxc always set errno.
* Rename "read_exact" macro in xc_domain_restore.c (which shadows
real function) to RDEXACT and change all callers.
* Make RDEXACT anamorphically use xch for error reporting rather than
* stderr.
* Call PERROR rather than ERROR when appropriate, so that log messages
include errno.
* Save errno in noncached_write so that its errno value is always
* right.
Signed-off-by: Ian Jackson <Ian.Jackson@eu.citrix.com>
oubuf and batch restore write messages should be sent with level
XTL_DEBUG so that they don't disturb progress output even with -v.
Signed-off-by: Ian Jackson <ian.jackson@eu.citrix.com>
This patch eliminate the global variables in libxenctrl (used for
logging and error reporting).
Instead the information which was in the global variables is now in a
new xc_interface* opaque structure, which xc_interface open returns
instead of the raw file descriptor; furthermore, logging is done via
xentoollog.
There are three new parameters to xc_interface_open to control the
logging, but existing callers can just pass "0" for all three to get
the old behaviour.
All libxc callers have been adjusted accordingly.
Also update QEMU_TAG for corresponding qemu change.
Signed-off-by: Ian Jackson <ian.jackson@eu.citrix.com>
Allow certain performance-critical hypercall wrappers to register data
buffers via a new interface which allows them to be 'bounced' into a
pre-mlock'ed page-sized per-thread data area. This saves the cost of
mlock/munlock on every such hypercall, which can be very expensive on
modern kernels.
Signed-off-by: Keir Fraser <keir.fraser@citrix.com>
To support wider than 28-bit MFNs, add XEN_DOMCTL_getpageframeinfo3
(with the type replacing the passed in MFN rather than getting or-ed
into it) to properly back xc_get_pfn_type_batch().
With xc_get_pfn_type_batch() only used internally to libxc, move its
prototype from xenctrl.h to xc_private.h.
This also fixes a couple of bugs in pre-existing code:
- the failure path for init_mem_info() leaked minfo->pfn_type,
- one error path of the XEN_DOMCTL_getpageframeinfo2 handler used
put_domain() where rcu_unlock_domain() was meant, and
- the XEN_DOMCTL_getpageframeinfo2 handler could call
xsm_getpageframeinfo() with an invalid struct page_info pointer.
Signed-off-by: Jan Beulich <jbeulich@novell.com>
Since the shared info layout is fixed, guests are required to use
VCPUOP_register_vcpu_info prior to booting any vCPU beyond the
traditional limit of 32.
MAX_VIRT_CPUS, being an implemetation detail of the hypervisor, is no
longer being exposed in the public headers.
The tools changes are clearly incomplete (and done only so things
would
build again), and the current state of the tools (using scalar
variables all over the place to represent vCPU bitmaps) very likely
doesn't permit booting DomU-s with more than the traditional number of
vCPU-s. Testing of the extended functionality was done with Dom0 (96
vCPU-s, as well as 128 vCPU-s out of which the kernel elected - by way
of a simple kernel side patch - to use only some, resulting in a
sparse
bitmap).
ia64 changes only to make things build, and build-tested only (and the
tools part only as far as the build would go without encountering
unrelated problems in the blktap code).
Signed-off-by: Jan Beulich <jbeulich@novell.com>
exact number of bytes. The consolidated versions are more watertight
than the various versions previously distributed around the library
source code.
Signed-off-by: Keir Fraser <keir@xensource.com>
across it. No longer make calls to the function dependent on 'non-live
save/restore': it's not a good test of whether the fd is a socket.
Signed-off-by: Keir Fraser <keir@xensource.com>
usage by means of fadvise64() to tell the OS to discard the cache
pages used for the save/dump file.
Signed-off-by: Simon Graham <Simon.Graham@stratus.com>
- An 'xc_error' struct is used to pass around error
details. Currently contains two members 'code' an enumeration of
error types, and 'message' a free text description of the specific
problem.
- The xc_get_last_error() method returns a const pointer to the
internal instance of this struct manged by libxc. By returning a
const pointer we can add extra members to the end of the struct at
any time without worrying about ABI of callers. This will let us
provide more fine-grained info if needed in the future.
- The xc_error instance is statically defined inside libxc and marked
__thread. This ensures that errors are recorded per-thread, and
that when dealing with errors we never need to call malloc - all
storage needed is statically allocated.
- The xc_clear_last_error() method resets any currently recorded
error details
- The xc_error_code_to_desc() method converts the integer error code
into a generic user facing messsage. eg "Invalid kernel". Together
with the 'message' field from xc_error, this provides the user
visible feedback. eg "Invalid kernel: Non PAE-kernel on PAE host."
- A callback can be registered with xc_set_error_handler to receive
notification whenever an error is recorded, rather than querying
for error details after the fact with xc_get_last_error
- If built with -DDEBUG set, a default error handler will be
registered which calls fprintf(stderr), thus maintaining current
behaviour of logging errors to stderr during developer builds.
- The python binding for libxc is updated to use xc_get_last_error
to pull out error details whenever appropriate, instead of
returning info based on 'errno'
- The xc_set_error method is private to libxc internals, and is used
for setting error details
- The ERROR and PERROR macros have been updated to call xc_set_error
automatically specifying XC_INTERNAL_ERROR as the error code. This
gives a generic error report for all current failure points
- Some uses of the ERROR macro have been replaced with explicit
calls to xc_set_error to enable finer grained error reporting. In
particular the code dealing with invalid kernel types uses this
to report about PAE/architecture/wordsize mismatches
The patch has been tested by calling xm create against a varietry of
config files defining invalid kernels of various kinds. It has also
been tested with libvirt talking to xend. In both cases the error
messages were propagated all the way back up the stack.
There is only one place where I need to do further work. The suspend
& restore APIs in Xend invoke external helper programs rather than
calling libxc directly. This means that error details are essentially
lost. Since there is already code in XenD which scans STDERR from
these programs I will investigate adapting this to extract actual
error messages from these helpers.
Signed-off-by: Daniel P. Berrange <berrange@redhat.com>
On solaris mlock requires a page aligned address and mlock doesn't
ensure the pages won't minor page fault.
Signed-off-by: Mark Johnson <mark.johnson@sun.com>
1. platform_op -- used by dom0 kernel to perform actions on the
hardware platform (e.g., MTRR access, microcode update, platform
quirks, ...)
2. domctl -- used by management tools to control a specified domain
3. sysctl -- used by management tools for system-wide actions
Benefits include more sensible factoring of actions to
hypercalls. Also allows tool compatibility to be tracked separately
from the dom0 kernel. The assumption is that it will be easier to
replace libxenctrl, libxenguest and Xen as a matched set if the
dom0 kernel does not need to be replaced too (e.g., because that
would require vendor revalidation).
From here on we hope to maintain dom0 kernel compatibility. This
promise is not extended to tool compatibility beyond the existing
guarantee that compatibility will not be broken within a three-level
stable release [3.0.2, 3.0.3, etc.].
Signed-off-by: Keir Fraser <keir@xensource.com>
Thus all can be disabled at compile time. It would be easy to
make enabling/disabling a run-time option too.
Signed-off-by: Keir Fraser <keir@xensource.com>
Make header path for kernel's privcmd/evtchn headers generic.
Remove pointless xi_*() interface that was using private libxc
interfaces.
Signed-off-by: John Levon <john.levon@sun.com>
move the shared info and grant table pfns into that block. This
allows us to remove the get_gnttablist dom0 op, and simplifies the
domain creation code slightly. Having the reserved block managed by
Xen may also make it slightly easier to handle the case where the
grant table needs to be extended at run time.
Suggested-by: kaf24
Signed-off-by: Steven Smith, sos22@cam.ac.uk
Ian Campbell
Ian Jackson
Christoph Egger
Keir Fraser
Keir Fraser
kfraser@localhost.localdomain
Ewan Mellor
kaf24@firebug.cl.cam.ac.uk
sos22@douglas.cl.cam.ac.uk