Andrew Cooper [Fri, 17 Aug 2012 13:46:49 +0000 (14:46 +0100)]
tools/python: Clean python correctly
Cleaning the python directory should completely remove the build/
directory, otherwise subsequent builds may be short-circuited and a
stale build installed.
Signed-off-by: Andrew Cooper <andrew.cooper3@citrix.com> Acked-by: Ian Campbell <ian.campbell@citrix.com> Committed-by: Ian Campbell <ian.campbell@citrix.com>
Wangzhenguo [Fri, 17 Aug 2012 13:46:48 +0000 (14:46 +0100)]
libxc/Linux: Add VM_DONTCOPY flag of the VMA of the hypercall buffer
This avoids the hypercall buffer becoming CoW on fork.
In multi-threads and multi-processes environment, e.g. the process has two
threads, thread A may call hypercall, thread B may call fork() to create child
process. After forking, all pages of the process including hypercall buffers
are cow. It will cause a write protection and return EFAULT error if hypervisor
calls copy_to_user in hypercall in thread A context,
Fix:
1. Before hypercall: use MADV_DONTFORK of madvise syscall to make the hypercall
buffer not to be copied to child process after fork.
2. After hypercall: undo the effect of MADV_DONTFORK for the hypercall buffer
by using MADV_DOFORK of madvise syscall.
3. Use mmap/nunmap for memory alloc/free instead of malloc/free to bypass libc.
Note:
Child processes must not use the opened xc_{interface,evtchn,gnttab,gntshr}
handle that inherits from parents. They should reopen the handle if they want
to interact with xc. Otherwise, it may cause segment fault to access hypercall
buffer caches of the handle.
Signed-off-by: Zhenguo Wang <wangzhenguo@huawei.com> Signed-off-by: Xiaowei Yang <xiaowei.yang@huawei.com> Acked-by: Ian Campbell <ian.campbell@citrix.com>
[ ijc -- s/ptr/p/ to fix build & tweaked the wording of the comments
slightly. ] Committed-by: Ian Campbell <ian.campbell@citrix.com>
M A Young [Fri, 17 Aug 2012 13:10:26 +0000 (14:10 +0100)]
xend: Replace the use of XMLPrettyPrint from PyXML with stdlib functionality.
This appears to have been missed by changeset 22235:b8cc53d22545
"Replace pyxml/xmlproc-based XML validator with lxml based one"
This was reported by Toshio Ernie Kuratomi at
https://bugzilla.redhat.com/show_bug.cgi?id=842843
Signed-off-by: Michael Young <m.a.young@durham.ac.uk> Acked-by: Ian Campbell <ian.campbell@citrix.com> Committed-by: Ian Campbell <ian.campbell@citrix.com>
Jan Beulich [Thu, 16 Aug 2012 16:38:05 +0000 (17:38 +0100)]
EPT/PoD: fix interaction with 1Gb pages
When PoD got enabled to support 1Gb pages, ept_get_entry() didn't get
updated to match - the assertion in there triggered, indicating that
the call to p2m_pod_demand_populate() needed adjustment.
Signed-off-by: Jan Beulich <jbeulich@suse.com> Acked-by: Tim Deegan <tim@xen.org> Committed-by: Tim Deegan <tim@xen.org>
Tim Deegan [Thu, 16 Aug 2012 13:31:09 +0000 (14:31 +0100)]
x86/mm: update max_mapped_pfn on MMIO mappings too.
max_mapped_pfn should reflect the highest mapping we've ever seen of
any type, or the tests in the lookup functions will be wrong. As it
happens, the highest mapping has always been a RAM one, but this is no
longer the case when we allow 64-bit BARs.
Reported-by: Xudong Hao <xudong.hao@intel.com> Signed-off-by: Tim Deegan <tim@xen.org> Committed-by: Tim Deegan <tim@xen.org>
Jan Beulich [Thu, 16 Aug 2012 08:16:19 +0000 (10:16 +0200)]
x86/PoD: clean up types
GMFN values must undoubtedly be "unsigned long". "count" and
"entry_count", since they are signed types, should also be "long" as
otherwise they can't fit all values that can fit into "d->tot_pages"
(which currently is "uint32_t").
Beyond that, the patch doesn't convert everything to "long" as in many
places it is clear that "int" suffices. In places where "long" is being
used partially already, the change is however being done.
Furthermore, page order values have no use of being "long".
Finally, in the course of updating a few printk messages anyway, some
also get slightly shortened (to focus on the relevant information).
Signed-off-by: Jan Beulich <jbeulich@suse.com> Acked-by: George Dunlap <george.dunlap@eu.citrix.com>
Jan Beulich [Thu, 16 Aug 2012 08:14:11 +0000 (10:14 +0200)]
x86/PoD: prevent guest from being destroyed upon early access to its memory
When an external agent (e.g. a monitoring daemon) happens to access the
memory of a PoD guest prior to setting the PoD target, that access must
fail for there not being any page in the PoD cache, and only the space
above the low 2Mb gets scanned for victim pages (while only the low 2Mb
got real pages populated so far).
To accomodate for this
- set the PoD target first
- do all physmap population in PoD mode (i.e. not just large [2Mb or
1Gb] pages)
- slightly lift the restrictions enforced by p2m_pod_set_mem_target()
to accomodate for the changed tools behavior
Tested-by: Jürgen Groß <juergen.gross@ts.fujitsu.com>
(in a 4.0.x based incarnation) Signed-off-by: Jan Beulich <jbeulich@suse.com> Acked-by: George Dunlap <george.dunlap@eu.citrix.com>
Boris Ostrovsky [Wed, 15 Aug 2012 07:43:25 +0000 (09:43 +0200)]
acpi: Make sure valid CPU is passed to do_pm_op()
Passing invalid CPU value to do_pm_op() will cause assertion
in cpu_online().
Signed-off-by: Boris Ostrovsky <boris.ostrovsky@amd.com>
Such checks would, at a first glance, then also be missing at the top
of various helper functions, but these check really were already
redundant with the check in do_pm_op(). Remove the redundant checks
for clarity and brevity.
Signed-off-by: Jan Beulich <jbeulich@suse.com> Committed-by: Jan Beulich <jbeulich@suse.com>
Daniel De Graaf [Wed, 15 Aug 2012 07:42:14 +0000 (09:42 +0200)]
x86-64/EFI: add CFLAGS to check compile
Without this, the compilation of check.c could fail due to compiler
features such as -fstack-protector being enabled, which causes a
missing __stack_chk_fail symbol error.
Signed-off-by: Daniel De Graaf <dgdegra@tycho.nsa.gov>
Rather than using plain CFLAGS here, remove CFLAGS-y from them to
particularly get rid of the -MF argument referencing (the undefined
here) $(@F).
The use of CFLAGS at once allows dropping the explicit use of -Werror.
Signed-off-by: Jan Beulich <jbeulich@suse.com> Committed-by: Jan Beulich <jbeulich@suse.com>
That c/s introduced a double unlock on the out-of-memory error path of
p2m_pod_demand_populate().
Signed-off-by: Jan Beulich <jbeulich@suse.com> Acked-by: George Dunlap <george.dunlap@eu.citrix.com> Acked-by: Andres Lagar-Cavilla <andres@lagarcavilla.org>
Andrew Cooper [Mon, 13 Aug 2012 17:09:33 +0000 (18:09 +0100)]
config: Split debug build from debug symbols
RPM based packaging systems expect binaries to have debug symbols which get
placed in a separate debuginfo RPM.
Split the concept of a debug build up so that binaries can be built with
debugging symbols without having the other gubbins which $(debug) implies, most
notibly frame pointers.
Signed-off-by: Andrew Cooper <andrew.cooper3@citrix.com> Tested-by: Ian Jackson <ian.jackson@eu.citrix.com> Committed-by: Ian Jackson <ian.jackson@eu.citrix.com>
Olaf Hering [Mon, 13 Aug 2012 13:11:22 +0000 (14:11 +0100)]
tools: init.d/Linux/xencommons: load all known backend drivers
Load all known backend drivers fron xenlinux and pvops based dom0
kernels. There is currently no code in xend or libxl to load these
drivers on demand. Currently libxl has also no helpful error message if
a backend driver is missing.
Signed-off-by: Olaf Hering <olaf@aepfle.de> Acked-by: Ian Jackson <ian.jackson@eu.citrix.com> Committed-by: Ian Jackson <ian.jackson@eu.citrix.com>
Ian Campbell [Tue, 7 Aug 2012 13:26:29 +0000 (14:26 +0100)]
libxl: write physical-device node if user did not supply a block script
This reverts one of the intentional changes from 25733:353bc0801b11.
That change exposed an issue with the xl migration protocol, which
although safe triggers the hotplug scripts device sharing logic.
For 4.2 we disable this logic by writing the physical-device xenstore
node ourselves if a user did not supply a script. If the user did
supply a script then we continue to rely on it to write the
physical-device node (not least because the script may create the
device and therefore it is not available before we run the script).
This means that to support localhost migration a block hotplug script
needs to be robust against adding a device twice and should not
deactivate the device until it has been removed twice.
This should be revisited for 4.3.
Signed-off-by: Ian Campbell <ian.campbell@citrix.com> Acked-by: Ian Jackson <Ian.Jackson@eu.citrix.com> Committed-by: Ian Campbell <ian.campbell@citrix.com>
Matt Wilson [Tue, 7 Aug 2012 06:49:53 +0000 (08:49 +0200)]
Although the "Intel Virtualization Technology FlexMigration
Application Note" (http://www.intel.com/Assets/PDF/manual/323850.pdf)
does not document support for extended model 2H model DH (Intel Xeon
Processor E5 Family), empirical evidence shows that the same MSR
addresses can be used for cpuid masking as exdended model 2H model AH
(Intel Xen Processor E3-1200 Family).
Signed-off-by: Matt Wilson <msw@amazon.com> Acked-by: Nakajima, Jun <jun.nakajima@intel.com> Committed-by: Jan Beulich <jbeulich@suse.com>
Ian Campbell [Mon, 6 Aug 2012 11:28:03 +0000 (12:28 +0100)]
libxl: support custom block hotplug scripts
These are provided using the "script=" syntax described in
docs/misc/xl-disk-configuration.txt.
The existing hotplug scripts currently conflate two different
concepts, namely that of making a datapath available in the backend
domain (logging into iSCSI LUNs and the like) and that of actually
connecting that datapath to a Xen backend path (e.g. writing
"physical-device" node in xenstore to bring up blkback).
For this reason the script support implemented here is only supported
in conjunction with backendtype=phy.
Eventually we hope to rework the hotplug scripts to separate the to
concepts, but that is not 4.2 material.
In addition there are some other subtleties:
- Previously in the blktap case we would add "script = .../blktap" to
the backend flex array, but then jumped to the PHY case which added
"script = .../block" too. The block one takes precendence since it
comes second.
This was, accidentally, correct. The blktap script is for blktap1
devices and not blktap2 devices. libxl completely manages the
blktap2 side of things without resorting to hotplug scripts and
creates a blkback device directly. Therefore the "block" script is
always the correct one to call. Custom script are not supported in
this context.
- libxl should not write the "physical-device" node. This is the
responsibility of the block script. Writing the "physical-device"
node in libxl basically completely short-cuts the standard block
hotplug script which uses "physical-device" to know if it has run
already or not.
In the case of more complex scripts libxl cannot know the right
value to write here anyway, in particular the device may not exist
until after the script is called.
This change has the side effect of re-enabling the checks for
device sharing aspect of the default block script, which I have tested
and which now cause libxl to properly abort now that libxl properly
checks for hotplug script errors.
There is no sharing check for blktap2 since even if you reuse the
same vhd the resulting tap device is different. I would have preferred
to simply write the "physical-device" node for the blktap2 case but
the hotplug script infrastructure is not currently setup to handle
LIBXL__DEVICE_KIND_VBD
devices without a hotplug script (backendtype phy and tap both end
up as KIND_VBD). Changing this was more surgery than I was happy doing
for 4.2 and therefore I have simply hardcoded to the block script for
the LIBXL_DISK_BACKEND_TAP case.
- libxl__device_disk_set_backend running against a phy device with a
script cannot stat the device to check its properties since it may
not exist until the script is run. Therefore I have special cased
this in disk_try_backend to simply assume that backend == phy is
always ok if a script was
configured. Similarly the other backend types are always rejected
if a script was configured.
Note that the reason for implementing the default script behaviour
in device_disk_add instead of libxl__device_disk_setdefault is
because we need to be able to tell when the script was
user-supplied rather than defaulted by libxl in order to correctly
implement the above. The setdefault function must be idempotent so
we cannot simply update disk->script.
I suspect that for 4.3 a script member should be added to
libxl__device, this would also help in the case above of handling
devices with no script in a consistent manner. This is not 4.2
material.
- When the block script falls through and shells out to a block-$type
script it used to pass "$node" however the only place this was
assigned was in the remove+phy case (in which case it contains the
file:// derived /dev/loopN device), and in that case the script
exits without falling through to the block-$type case.
Since libxl never creates a type other than phy this never happens
in practice anyway and we now call the correct block-$type script
directly. But fix it up anyway since it is confusing.
- The block-nbd and block-enbd scripts which we supply appear to be
broken WRT the hotplug calling convention, in that they seem to
expect a command line parameter (perhaps the $node described above)
rather than reading the appropriate node from xenstore.
I rather suspect this was broken by 7774:e2e7f47e6f79 in November
2005. I think it is safe to say no one is using these scripts! I
haven't fixed this here. It would be good to track down some working
scripts and either incorproate them or defer to them in their existing
home (e.g. if they live somewhere useful like the nbd tools
package).
- Added a few block script related entries to check-xl-disk-parse
from http://backdrift.org/xen-block-iscsi-script-with-multipath-support
and http://lists.linbit.com/pipermail/drbd-user/2008-September/010221.html /
http://www.drbd.org/users-guide-emb/s-xen-configure-domu.html (and
snuck in another interesting empty CDROM case)
This highlighted two bugs in the libxlu disk parser handling of the
deprecated "<script>:" prefix:
- It was failing to prefix with "block-" to construct the actual
script name
- The regex for matching iscsi or drdb or e?nbd was incorrect
- Use libxl__abs_path for the nic script too. Just because the
existing code nearly tricked me into repeating the mistake
I have tested with a custom block script which uses "lvchange -a" to
dynamically add remove the referenced device (simulates iSCSI
login/logout without requiring me to faff around setting up an iSCSI
target). I also tested on a blktap2 system.
I haven't directly tested anything more complex like iscsi: or nbd:
other than what check-xl-disk-parse exercises.
[ Recommit of correct version of 25727:a8d708fcb347, which was mangled
during commit. Sorry. -iwj ]
Signed-off-by: Ian Campbell <ian.campbell@citrix.com> Acked-by: Ian Jackson <ian.jackson@eu.citrix.com> Committed-by: Ian Jackson <ian.jackson@eu.citrix.com>
Jan Beulich [Mon, 6 Aug 2012 09:18:43 +0000 (11:18 +0200)]
x86: fix wait code asm() constraints
This fixes theoretical issues with those constraints - operands that
get clobbered before consuming all input operands must be marked so
according the the gcc documentation. Beyond that, the change is merely
code improvement, not a bug fix.
In __prepare_to_wait(), properly mark early clobbered registers. By
doing so, we at once eliminate the need to save/restore rCX and rDI.
In check_wakeup_from_wait(), make the current constraints match by
removing the code that actuall alters registers. By adjusting the
resume address in __prepare_to_wait(), we can simply re-use the copying
operation there (rather than doing a second pointless copy in the
opposite direction after branching to the resume point), which at once
eliminates the need for re-loading rCX and rDI inside the asm().
Signed-off-by: Jan Beulich <jbeulich@suse.com> Acked-by: Keir Fraser <keir@xen.org>
Ian Campbell [Fri, 3 Aug 2012 11:25:29 +0000 (12:25 +0100)]
libxl: fix cleanup of tap devices in libxl__device_destroy
We pass be_path to tapdisk_destroy but we've already deleted it so it
fails to read tapdisk-params. However it appears that we need to
destroy the tap device after tearing down xenstore, to avoid the leak
reported by Greg Wettstein in
<201207312141.q6VLfJje012656@wind.enjellic.com>.
So read the tapdisk-params in the cleanup transaction, before the
remove, and pass that down to destroy_tapdisk instead. tapdisk-params
may of course be NULL if the device isn't a tap device.
There is no need to tear down the tap device from
libxl__initiate_device_remove since this ultimately calls
libxl__device_destroy.
Propagate and log errors from libxl__device_destroy_tapdisk.
Signed-off-by: Ian Campbell <ian.campbell@citrix.com> Acked-by: Ian Jackson <ian.jackson@eu.citrix.com> Committed-by: Ian Jackson <ian.jackson@eu.citrix.com>
Ian Campbell [Fri, 3 Aug 2012 10:59:12 +0000 (11:59 +0100)]
libxl: support custom block hotplug scripts
These are provided using the "script=" syntax described in
docs/misc/xl-disk-configuration.txt.
The existing hotplug scripts currently conflate two different
concepts, namely that of making a datapath available in the backend
domain (logging into iSCSI LUNs and the like) and that of actually
connecting that datapath to a Xen backend path (e.g. writing
"physical-device" node in xenstore to bring up blkback).
For this reason the script support implemented here is only supported
in conjunction with backendtype=phy.
Eventually we hope to rework the hotplug scripts to separate the to
concepts, but that is not 4.2 material.
In addition there are some other subtleties:
- Previously in the blktap case we would add "script = .../blktap" to
the backend flex array, but then jumped to the PHY case which added
"script = .../block" too. The block one takes precendence since it
comes second.
This was, accidentally, correct. The blktap script is for blktap1
devices and not blktap2 devices. libxl completely manages the
blktap2 side of things without resorting to hotplug scripts and
creates a blkback device directly. Therefore the "block" script is
always the correct one to call. Custom script are not supported in
this context.
- libxl should not write the "physical-device" node. This is the
responsibility of the block script. Writing the "physical-device"
node in libxl basically completely short-cuts the standard block
hotplug script which uses "physical-device" to know if it has run
already or not.
In the case of more complex scripts libxl cannot know the right
value to write here anyway, in particular the device may not exist
until after the script is called.
This change has the side effect of re-enabling the checks for
device sharing aspect of the default block script, which I have tested
and which now cause libxl to properly abort now that libxl properly
checks for hotplug script errors.
There is no sharing check for blktap2 since even if you reuse the
same vhd the resulting tap device is different. I would have preferred
to simply write the "physical-device" node for the blktap2 case but
the hotplug script infrastructure is not currently setup to handle
LIBXL__DEVICE_KIND_VBD
devices without a hotplug script (backendtype phy and tap both end
up as KIND_VBD). Changing this was more surgery than I was happy doing
for 4.2 and therefore I have simply hardcoded to the block script for
the LIBXL_DISK_BACKEND_TAP case.
- libxl__device_disk_set_backend running against a phy device with a
script cannot stat the device to check its properties since it may
not exist until the script is run. Therefore I have special cased
this in disk_try_backend to simply assume that backend == phy is
always ok if a script was
configured. Similarly the other backend types are always rejected
if a script was configured.
Note that the reason for implementing the default script behaviour
in device_disk_add instead of libxl__device_disk_setdefault is
because we need to be able to tell when the script was
user-supplied rather than defaulted by libxl in order to correctly
implement the above. The setdefault function must be idempotent so
we cannot simply update disk->script.
I suspect that for 4.3 a script member should be added to
libxl__device, this would also help in the case above of handling
devices with no script in a consistent manner. This is not 4.2
material.
- When the block script falls through and shells out to a block-$type
script it used to pass "$node" however the only place this was
assigned was in the remove+phy case (in which case it contains the
file:// derived /dev/loopN device), and in that case the script
exits without falling through to the block-$type case.
Since libxl never creates a type other than phy this never happens
in practice anyway and we now call the correct block-$type script
directly. But fix it up anyway since it is confusing.
- The block-nbd and block-enbd scripts which we supply appear to be
broken WRT the hotplug calling convention, in that they seem to
expect a command line parameter (perhaps the $node described above)
rather than reading the appropriate node from xenstore.
I rather suspect this was broken by 7774:e2e7f47e6f79 in November
2005. I think it is safe to say no one is using these scripts! I
haven't fixed this here. It would be good to track down some working
scripts and either incorproate them or defer to them in their existing
home (e.g. if they live somewhere useful like the nbd tools
package).
- Added a few block script related entries to check-xl-disk-parse
from http://backdrift.org/xen-block-iscsi-script-with-multipath-support
and http://lists.linbit.com/pipermail/drbd-user/2008-September/010221.html /
http://www.drbd.org/users-guide-emb/s-xen-configure-domu.html (and
snuck in another interesting empty CDROM case)
This highlighted two bugs in the libxlu disk parser handling of the
deprecated "<script>:" prefix:
- It was failing to prefix with "block-" to construct the actual
script name
- The regex for matching iscsi or drdb or e?nbd was incorrect
- Use libxl__abs_path for the nic script too. Just because the
existing code nearly tricked me into repeating the mistake
I have tested with a custom block script which uses "lvchange -a" to
dynamically add remove the referenced device (simulates iSCSI
login/logout without requiring me to faff around setting up an iSCSI
target). I also tested on a blktap2 system.
I haven't directly tested anything more complex like iscsi: or nbd:
other than what check-xl-disk-parse exercises.
[ reran flex/bison -iwj ]
Signed-off-by: Ian Campbell <ian.campbell@citrix.com> Acked-by: Ian Jackson <ian.jackson@eu.citrix.com>
Ian Jackson [Fri, 3 Aug 2012 10:57:10 +0000 (11:57 +0100)]
libxl: correct some comments regarding event API and fds
* libxl may indeed register more than one callback for the same fd,
with some restrictions. The allowable range of responses to this by
the application means that this should pose no problems for users.
But the documentation comment should be fixed.
* Document the relaxed synchronicity semantics of the fd_modify
registration callback.
* A couple of comments referred to old names for functions.
Signed-off-by: Ian Jackson <ian.jackson@eu.citrix.com> Acked-by: Ian Campbell <ian.campbell@citrix.com>
Ian Campbell [Fri, 3 Aug 2012 08:54:17 +0000 (09:54 +0100)]
nestedhvm: fix nested page fault build error on 32-bit
cc1: warnings being treated as errors
hvm.c: In function ‘hvm_hap_nested_page_fault’:
hvm.c:1282: error: passing argument 2 of ‘nestedhvm_hap_nested_page_fault’ from incompatible pointer type /local/scratch/ianc/devel/xen-unstable.hg/xen/include/asm/hvm/nestedhvm.h:55: note: expected ‘paddr_t *’ but argument is of type ‘long unsigned int *’
hvm_hap_nested_page_fault takes an unsigned long gpa and passes &gpa
to nestedhvm_hap_nested_page_fault which takes a paddr_t *. Since both
of the callers of hvm_hap_nested_page_fault (svm_do_nested_pgfault and
ept_handle_violation) actually have the gpa which they pass to
hvm_hap_nested_page_fault as a paddr_t I think it makes sense to
change the argument to hvm_hap_nested_page_fault.
The other user of gpa in hvm_hap_nested_page_fault is a call to
p2m_mem_access_check, which currently also takes a paddr_t gpa but I
think a paddr_t is appropriate there too.
Jan points out that this is also an issue for >4GB guests on the 32
bit hypervisor.
Signed-off-by: Ian Campbell <ian.campbell@citrix.com> Acked-by: Jan Beulich <jbeulich@suse.com> Acked-by: Tim Deegan <tim@xen.org> Committed-by: Ian Campbell <ian.campbell@citrix.com>
Ian Campbell [Fri, 3 Aug 2012 08:54:16 +0000 (09:54 +0100)]
libxl: const correctness for libxl__xs_path_cleanup
Signed-off-by: Ian Campbell <ian.campbell@citrix.com> Acked-by: Ian Jackson <ian.jackson@eu.citrix.com> Committed-by: Ian Campbell <ian.campbell@citrix.com>
Ian Jackson [Fri, 3 Aug 2012 08:54:15 +0000 (09:54 +0100)]
libxl: add a comment re the memory management API instability
Signed-off-by: Ian Jackson <ian.jackson@eu.citrix.com> Acked-by: Ian Campbell <ian.campbell@citrix.com> Committed-by: Ian Campbell <ian.campbell@citrix.com>
However I think it is wrong that this function is unused, this and
libxl_event_init should be used by libxl__event_new. As it happens
both are just memset to zero but for correctness we should use the
init functions (in case the IDL changes).
In the generator we also need to properly handle init_var == 0 which
the current if statements incorrectly treat as False. This doesn't
actually have any impact on the generated code.
Signed-off-by: Ian Campbell <ian.campbell@citrix.com> Acked-by: Ian Jackson <ian.jackson@eu.citrix.com> Committed-by: Ian Campbell <ian.campbell@citrix.com>
Ian Jackson [Fri, 3 Aug 2012 08:54:14 +0000 (09:54 +0100)]
libxl: remove an unused numainfo parameter
Signed-off-by: Ian Jackson <ian.jackson@eu.citrix.com> Acked-by: Ian Campbell <ian.campbell@citrix.com> Acked-by: Dario Faggioli <dario.faggioli@citrix.com> Committed-by: Ian Campbell <ian.campbell@citrix.com>
Ian Jackson [Fri, 3 Aug 2012 08:54:13 +0000 (09:54 +0100)]
libxl: remus: mark TODOs more clearly
Change the TODOs in the remus code to "REMUS TODO" which will make
them easier to grep for later. AIUI all of these are essential for
use of remus in production.
Also add a new TODO and a new assert, to check rc on entry to
remus_checkpoint_dm_saved.
Signed-off-by: Ian Jackson <ian.jackson@eu.citrix.com> Acked-by: Ian Campbell <ian.campbell@citrix.com> Committed-by: Ian Campbell <ian.campbell@citrix.com>
Ian Jackson [Fri, 3 Aug 2012 08:54:12 +0000 (09:54 +0100)]
libxl: do not blunder on if bootloader fails (again)
Do not lose the rc value passed to bootloader_callback. Do not lose
the rc value from the bl when the local disk detach succeeds.
While we're here rationalise the use of bl->rc to make things clearer.
Set it to zero at the start and always update it conditionally; copy
it into bootloader_callback's argument each time.
Signed-off-by: Ian Jackson <ian.jackson@eu.citrix.com> Acked-by: Ian Campbell <ian.campbell@citrix.com> Committed-by: Ian Campbell <ian.campbell@citrix.com>
Ian Jackson [Fri, 3 Aug 2012 08:54:11 +0000 (09:54 +0100)]
libxl: rename aodevs to multidev
To be consistent with the new function naming, rename
libxl__ao_devices to libxl__multidev and all variables aodevs to
multidev.
No functional change.
Signed-off-by: Ian Jackson <ian.jackson@eu.citrix.com> Acked-by: Ian Campbell <ian.campbell@citrix.com> Committed-by: Ian Campbell <ian.campbell@citrix.com>
Ian Jackson [Fri, 3 Aug 2012 08:54:11 +0000 (09:54 +0100)]
libxl: abolish useless `start' parameter to libxl__add_*
0 is always passed for this parameter and the code doesn't, actually,
use it, now.
Signed-off-by: Ian Jackson <ian.jackson@eu.citrix.com> Acked-by: Ian Campbell <ian.campbell@citrix.com> Committed-by: Ian Campbell <ian.campbell@citrix.com>
Ian Jackson [Fri, 3 Aug 2012 08:54:10 +0000 (09:54 +0100)]
libxl: fix formatting of DEFINE_DEVICES_ADD
These lines were exactly 80 columns wide, which produces hideous wrap
damage in an 80 column emacs. Reformat using emacs's C-c \,
which puts the \ in column 72 (by default) where possible.
Whitespace change only.
Signed-off-by: Ian Jackson <ian.jackson@eu.citrix.com> Acked-by: Ian Campbell <ian.campbell@citrix.com> Committed-by: Ian Campbell <ian.campbell@citrix.com>
Ian Jackson [Fri, 3 Aug 2012 08:54:09 +0000 (09:54 +0100)]
libxl: fix device counting race in libxl__devices_destroy
Don't have a fixed number of devices in the aodevs array, and instead
size it depending on the devices present in xenstore. Somewhat
formalise the multiple device addition/removal machinery to make this
clearer and easier to do.
As a side-effect we fix a few "lost thread of control" bug which would
occur if there were no devices of a particular kind. (Various if
statements which checked for there being no devices have become
redundant, but are retained to avoid making the patch bigger.)
Specifically:
* Users of libxl__ao_devices are no longer expected to know in
advance how many device operations they are going to do. Instead
they can initiate them one at a time, between bracketing calls to
"begin" and "prepared".
* The array of aodevs used for this is dynamically sized; to support
this it's an array of pointers rather than of structs.
* Users of libxl__ao_devices are presented with a more opaque interface.
They are are no longer expected to, themselves,
- look into the array of aodevs (this is now private)
- know that the individual addition/removal completions are
handled by libxl__ao_devices_callback (this callback function
is now a private function for the multidev machinery)
- ever deal with populating the contents of an aodevs
* The doc comments relating to some of the members of
libxl__ao_device are clarified. (And the member `aodevs' is moved
to put it with the other members with the same status.)
* The multidev machinery allocates an aodev to represent the
operation of preparing all of the other operations. See
the comment in libxl__multidev_begin.
A wrinkle is that the functions are called "multidev" but the structs
are called "libxl__ao_devices" and "aodevs". I have given these
functions this name to distinguish them from "libxl__ao_device" and
"aodev" and so forth by more than just the use of the plural "s"
suffix.
In the next patch we will rename the structs.
Signed-off-by: Roger Pau Monne <roger.pau@citrix.com> Signed-off-by: Ian Jackson <ian.jackson@eu.citrix.com> Acked-by: Ian Campbell <ian.campbell@eu.citrix.com> Committed-by: Ian Campbell <ian.campbell@citrix.com>
Ian Jackson [Fri, 3 Aug 2012 08:54:08 +0000 (09:54 +0100)]
libxl: react correctly to bootloader pty master POLLHUP
Receive POLLHUP on the bootloader master pty is not an error.
Hopefully it means that the bootloader has exited and therefore the
pty slave side has no process group any more. (At least NetBSD
indicates POLLHUP on the master in this case.)
So send the bootloader SIGTERM; if it has already exited then this has
no effect (except that on some versions of NetBSD it erroneously
returns ESRCH and we print a harmless warning) and we will then
collect the bootloader's exit status and be satisfied.
However, we remember that we have done this so that if we got POLLHUP
for some other reason than that the bootloader exited we report
something resembling a useful message.
In order to implement this we need to provide a way for users of
datacopier to handle POLLHUP rather than treating it as fatal.
We rename bootloader_abort to bootloader_stop since it now no longer
only applies to error situations.
Signed-off-by: Roger Pau Monne <roger.pau@citrix.com> Signed-off-by: Ian Jackson <ian.jackson@eu.citrix.com> Acked-by: Ian Campbell <ian.campbell@citrix.com> Committed-by: Ian Campbell <ian.campbell@citrix.com>
Ian Jackson [Fri, 3 Aug 2012 08:54:07 +0000 (09:54 +0100)]
libxl: unify libxl__device_destroy and device_hotplug_done
device_hotplug_done contains an open-coded but improved version of
libxl__device_destroy. So move the contents of device_hotplug_done
into libxl__device_destroy, deleting the old code, and replace it at
its old location with a function call.
Add the missing call to libxl__xs_transaction_abort (which was present
in neither version and technically speaking is always a no-op with
this code as it stands at the moment because no-one does "goto out"
other than after libxl__xs_transaction_start or _commit).
Also fix the error handling: the rc from the destroy should be
propagated into the aodev.
Reported-by: Ian Campbell <Ian.Campbell@citrix.com> Signed-off-by: Ian Jackson <ian.jackson@eu.citrix.com> Acked-by: Ian Campbell <ian.campbell@citrix.com> Committed-by: Ian Campbell <ian.campbell@citrix.com>
Olaf Hering [Fri, 3 Aug 2012 08:54:07 +0000 (09:54 +0100)]
tools/vtpm: fix tpm_version.h error during parallel build
Generating the tpm_version.h is not make -j safe:
In file included from ../tpm/tpm_emulator.h:25:0,
from ../tpm/tpm_startup.c:18:
../tpm/tpm_version.h:1:0: error: unterminated #ifndef
make[5]: *** [tpm_startup.o] Error 1
This happens because make can not know that 'all-recursive' depends on
'version'. Fix this by calling the individual make targets. Doing it
this way avoids adding yet another patch to the downloaded source.
Signed-off-by: Olaf Hering <olaf@aepfle.de> Acked-by: Ian Campbell <ian.campbell@citrix.com> Committed-by: Ian Campbell <ian.campbell@citrix.com>
Ian Campbell [Fri, 3 Aug 2012 08:54:06 +0000 (09:54 +0100)]
libxl: only read script once in libxl__hotplug_*
instead of duplicating the error handling etc in get_hotplug_env just pass the
script already read by the caller down.
Signed-off-by: Ian Campbell <ian.campbell@citrix.com> Acked-by: Ian Jackson <ian.jackson@eu.citrix.com> Committed-by: Ian Campbell <ian.campbell@citrix.com>
Ian Campbell [Fri, 3 Aug 2012 08:54:05 +0000 (09:54 +0100)]
libxl: prefix *.for-check with _ to mark it as a generated file.
Keeps it out of my greps etc.
Signed-off-by: Ian Campbell <ian.campbell@citrix.com> Acked-by: Ian Jackson <ian.jackson@eu.citrix.com> Committed-by: Ian Campbell <ian.campbell@citrix.com>
Ian Campbell [Fri, 3 Aug 2012 08:54:03 +0000 (09:54 +0100)]
libxl: make libxl_device_pci_{add,remove,destroy} interfaces asynchronous
This does not make the implementation fully asynchronous but just
updates the API to support asynchrony in the future.
Currently although these functions do not call hotplug scripts etc and
therefore are not "slow" (per the comment about ao machinery in
libxl_internal.h) they do interact with the device model and so are
not quite "fast" either. We can live with this for now.
Signed-off-by: Ian Campbell <ian.campbell@citrix.com> Acked-by: Ian Jackson <ian.jackson@eu.citrix.com> Committed-by: Ian Campbell <ian.campbell@citrix.com>
David Vrabel [Fri, 3 Aug 2012 07:50:28 +0000 (09:50 +0200)]
cpufreq: P state stats aren't available if there is no cpufreq driver
If there is no cpufreq driver (e.g., with an AMD Opteron 8212) then
reading the P state statistics causes a deadlock as an uninitialized
spinlock is locked in do_get_pm_info(). The spinlock is initialized in
cpufreq_statistic_init() which is not called if cpufreq_driver == NULL.
Signed-off-by: David Vrabel <david.vrabel@citrix.com> Committed-by: Jan Beulich <jbeulich@suse.com>
Tim Deegan [Thu, 2 Aug 2012 13:44:53 +0000 (14:44 +0100)]
nestedhvm: return the pfec from the pagetable walk.
Don't clobber it with the pfec from teh p2m walk behind it; the guest
will not expect (or be able to handle) error codes that come from the
p2m table, which it can't see or control.
Signed-off-by: Tim Deegan <tim@xen.org> Acked-by: Christoph Egger <Christoph.Egger@amd.com> Committed-by: Tim Deegan <tim@xen.org>
Christoph Egger [Thu, 2 Aug 2012 13:38:09 +0000 (14:38 +0100)]
nestedhvm: fix write access fault on ro mapping
Fix write access fault when host npt is mapped read-only.
In this case let the host handle the #NPF.
Apply host p2mt to hap-on-hap pagetable entry.
This fixes the l2 guest graphic display refresh problem.
Signed-off-by: Christoph Egger <Christoph.Egger@amd.com> Acked-by: Tim Deegan <tim@xen.org> Committed-by: Tim Deegan <tim@xen.org>
Ian Jackson [Wed, 1 Aug 2012 11:46:52 +0000 (12:46 +0100)]
libxl: enforce prohibitions of internal callers
libxl_internal.h says:
* Functions using LIBXL__INIT_EGC may *not* generally be called from
* within libxl, because libxl__egc_cleanup may call back into the
* application. ...
and
* ... [Functions which take an ao_how] MAY NOT
* be called from inside libxl, because they can cause reentrancy
* callbacks.
However, this was not enforced. Particularly the latter restriction
is easy to overlook, especially since during the transition period to
the new event system we have bent this rule a couple of times, and the
bad pattern simply involves passing 0 or NULL for the ao_how.
So use the compiler to enforce this property, as follows:
- Mark all functions which take a libxl_asyncop_how, or which
use EGC_INIT or LIBXL__INIT_EGC, with a new annotation
LIBXL_EXTERNAL_CALLERS_ONLY in the public header.
- Change the documentation comment for asynch operations and egcs to
say that this should always be done.
- Arrange that if libxl.h is included via libxl_internal.h,
LIBXL_EXTERNAL_CALLERS_ONLY expands to __attribute__((warning(...))),
which generates a message like this:
libxl.c:1772: warning: call to 'libxl_device_disk_remove'
declared with attribute warning:
may not be called from within libxl
Otherwise, the annotation expands to nothing, so external
callers are unaffected.
- Forbid inclusion of both libxl.h and libxl_internal.h unless
libxl_internal.h came first, so that the above check doesn't have
any loopholes. Files which include libxl_internal.h should not
include libxl.h as well.
This is enforced explicitly using #error. However, in practice
with the current tree it just changes the error message when this
mistake is made; otherwise we would carry on to immediately
following #define which would cause the compiler to complain that
LIBXL_EXTERNAL_CALLERS_ONLY was redefined. Then the developer
might be tempted to add a #ifndef which would be wrong - it would
leave the affected translation unit unprotected by the new
enforcement regime. So let's be explicit.
- Fix the one source of files which violate the above principle, the
output from the idl compiler, by removing the redundant inclusion
of libxl.h from the output.
Also introduce a new script "check-libxl-api-rules" which contains
some ad-hoc regexps to spot and complain when libxl.h contains
functions which mention libxl_asyncop_how but not
LIBXL_EXTERNAL_CALLERS_ONLY. This isn't a full C parser but is likely
to get the common cases right and err on the side of complaining.
While we are here, the invocation of perl for the bsd queue.h seddery
to $(PERL).
Signed-off-by: Ian Jackson <ian.jackson@eu.citrix.com> Cc: Roger Pau Monne <roger.pau@citrix.com> Acked-by: Ian Campbell <ian.campbell@citrix.com> Committed-by: Ian Campbell <ian.campbell@citrix.com>
Andrew Cooper [Wed, 1 Aug 2012 11:46:51 +0000 (12:46 +0100)]
tools/config: Allow building of components to be controlled from .config
For build systems which build certain Xen components separately, allow certain
components to be conditionally built based on .config, rather than always
building them.
This patch allows qemu and blktap to be configured in this manner.
Signed-off-by: Andrew Cooper <andrew.cooper3@citrix.com> Acked-by: Ian Jackson <ian.jackson@eu.citrix.com> Committed-by: Ian Campbell <ian.campbell@citrix.com>
Andrew Cooper [Wed, 1 Aug 2012 11:46:50 +0000 (12:46 +0100)]
tools/ocaml: ignore and clean .spot and .spit files
Newer ocaml toolchains generate .spot and .spit files which are ocaml metadata
about their respective source files.
Add them to the clean rules as well as the .{hg,git}ignore files.
Signed-off-by: Andrew Cooper <andrew.cooper3@citrix.com> Acked-by: Ian Jackson <ian.jackson@eu.citrix.com> Committed-by: Ian Campbell <ian.campbell@citrix.com>
Olaf Hering [Wed, 1 Aug 2012 11:46:49 +0000 (12:46 +0100)]
libxl: fix unitialized variables in libxl__primary_console_find
gcc 4.5 as shipped with openSuSE 11.4 does not recognize the case of
LIBXL_DOMAIN_TYPE_INVALID properly:
cc1: warnings being treated as errors
libxl.c: In function 'libxl_primary_console_exec':
libxl.c:1408:14: error: 'domid' may be used uninitialized in this function
libxl.c:1409:9: error: 'cons_num' may be used uninitialized in this function
libxl.c:1410:24: error: 'type' may be used uninitialized in this function
libxl.c: In function 'libxl_primary_console_get_tty':
libxl.c:1421:14: error: 'domid' may be used uninitialized in this function
libxl.c:1422:9: error: 'cons_num' may be used uninitialized in this function
libxl.c:1423:24: error: 'type' may be used uninitialized in this function
make[3]: *** [libxl.o] Error 1
Fix this by adding a default case.
Signed-off-by: Olaf Hering <olaf@aepfle.de> Acked-by: Ian Campbell <ian.campbell@citrix.com> Committed-by: Ian Campbell <ian.campbell@citrix.com>
Roger Pau Monne [Wed, 1 Aug 2012 11:46:48 +0000 (12:46 +0100)]
init/NetBSD: move xenbackendd to xend init script
xenbackendd is not needed by the xl toolstack, so move it's launch to
the xend script.
We have to iterate until we are sure there are no xend processes left,
since doing a single pkill usually leaves xend processes running.
Signed-off-by: Roger Pau Monne <roger.pau@citrix.com> Acked-by: Christoph Egger <Christoph.Egger@amd.com> Acked-by: Ian Jackson <ian.jackson@eu.citrix.com> Tested-by: Christoph Egger <Christoph.Egger@amd.com> Committed-by: Ian Campbell <ian.campbell@citrix.com>
Roger Pau Monne [Wed, 1 Aug 2012 11:46:47 +0000 (12:46 +0100)]
libxl: call hotplug scripts from xl for NetBSD
Add the missing NetBSD functions to call hotplug scripts, and disable
xenbackendd if libxl/disable_udev is not set.
Signed-off-by: Roger Pau Monne <roger.pau@citrix.com> Acked-by: Ian Jackson <ian.jackson@eu.citrix.com> Acked-by: Christoph Egger <Christoph.Egger@amd.com> Committed-by: Ian Campbell <ian.campbell@citrix.com>
Roger Pau Monne [Wed, 1 Aug 2012 11:46:47 +0000 (12:46 +0100)]
tools/build: fix pygrub linking
Prevent creating a symlink to $(DESTDIR)/$(BINDIR) if it is the same
as $(PRIVATE_BINDIR)
This fixes NetBSD install, where $(DESTDIR)/$(BINDIR) ==
$(PRIVATE_BINDIR).
Cc: Ian Jackson <ian.jackson@eu.citrix.com> Cc: Christoph Egger <Christoph.Egger@amd.com> Signed-off-by: Roger Pau Monne <roger.pau@citrix.com> Acked-by: Ian Campbell <ian.campbell@citrix.com> Committed-by: Ian Campbell <ian.campbell@citrix.com>
Jan Beulich [Fri, 27 Jul 2012 08:00:12 +0000 (10:00 +0200)]
x86-64: drop updating of UREGS_rip when converting sysenter to #GP
This was set to zero immediately before the #GP injection code, since
SYSENTER doesn't really have a return address.
Reported-by: Ian Campbell <Ian.Campbell@citrix.com>
Furthermore, UREGS_cs and UREGS_rip don't need to be written a second
time, as the PUSHes above already can/do take care of putting in place
the intended values.
Signed-off-by: Jan Beulich <jbeulich@suse.com> Acked-by: Ian Campbell <ian.campbell@citrix.com> Acked-by: Keir Fraser <keir@xen.org>
Ian Jackson [Thu, 26 Jul 2012 16:22:39 +0000 (17:22 +0100)]
libxl: fix reentrancy hazard in fd event processing
In afterpoll_internal, the callback functions may register and
deregister events arbitrarily. This means that we need to consider
the reentrancy-safety of the event machinery state variables.
Most of the code is safe but the fd handling is not. Fix this by
arranging to restart the fd scan loop every time we call one of these
callback functions.
For this loop to terminate, we modify afterpoll_check_fd so that it
returns only once for each of afterpoll's efds.
Another possible solution would be simply to return from
afterpoll_internal after calling efd->func. That would be a small and
more obviously correct change but would prevent the process from
handling more than one fd event with a single call to poll.
This is apropos of a report from Roger Pau Monne to me (pers.comm.)
of this crash on NetBSD:
Program terminated with signal 11, Segmentation fault.
#0 0x00007f7ff743131b in afterpoll_check_fd (poller=<optimized out>, fds=0x7f7ff7b241c0, nfds=7, fd=-1, events=1)
at libxl_event.c:856
856 if (fds[slot].fd != fd)
Signed-off-by: Ian Jackson <ian.jackson@eu.citrix.com> Reported-by: Roger Pau Monne <roger.pau@citrix.com> Tested-by: Roger Pau Monne <roger.pau@citrix.com> Acked-by: Ian Campbell <Ian.Campbell@citrix.com> Committed-by: Ian Campbell <ian.campbell@citrix.com>
Jan Beulich [Thu, 26 Jul 2012 15:51:51 +0000 (16:51 +0100)]
x86/hvm: don't leave emulator in inconsistent state
The fact that handle_mmio(), and thus the instruction emulator, is
being run through twice for emulations that require involvement of the
device model, allows for the second run to see a different guest state
than the first one. Since only the MMIO-specific emulation routines
update the vCPU's io_state, if they get invoked on the second pass,
internal state (and particularly this variable) can be left in a state
making successful emulation of a subsequent MMIO operation impossible.
Consequently, whenever the emulator invocation returns without
requesting a retry of the guest instruction, reset io_state.
Signed-off-by: Jan Beulich <jbeulich@suse.com> Acked-by: Keir Fraser <keir@xen.org> Committed-by: Ian Jackson <ian.jackson@eu.citrix.com>
libxl: convert libxl_device_vfb_add to an async operation
Split libxl_device_vfb_add into libxl__device_vfb_add (to be used
inside already running ao's), and make libxl_device_vfb_add a stub
to call libxl__device_vfb_add.
Signed-off-by: Roger Pau Monne <roger.pau@citrix.com> Acked-by: Ian Jackson <ian.jackson@eu.citrix.com> Committed-by: Ian Campbell <ian.campbell@citrix.com>
libxl: convert libxl_device_vkb_add to an async operation
Split libxl_device_vkb_add into libxl__device_vkb_add (to be used
inside already running ao's), and make libxl_device_vkb_add a stub to
call libxl__device_vkb_add.
Signed-off-by: Roger Pau Monne <roger.pau@citrix.com> Acked-by: Ian Jackson <ian.jackson@eu.citrix.com> Committed-by: Ian Campbell <ian.campbell@citrix.com>
libxl: call hotplug scripts for nic devices from libxl
Since most of the needed work is already done in previous patches,
this patch only contains the necessary code to call hotplug scripts
for nic devices, that should be called when the device is added or
removed from a guest.
Added another parameter to libxl__get_hotplug_script_info, that is
used to know the number of times hotplug scripts have been called for
that device. This is currently used by IOEMU nics on Linux.
Signed-off-by: Roger Pau Monne <roger.pau@citrix.com> Acked-by: Ian Jackson<ian.jackson@eu.citrix.com> Committed-by: Ian Campbell <ian.campbell@citrix.com>
libxl: call hotplug scripts for disk devices from libxl
Since most of the needed work is already done in previous patches,
this patch only contains the necessary code to call hotplug scripts
for disk devices, that should be called when the device is added or
removed from a guest.
We will chain the launch of the disk hotplug scripts after the
device_backend_callback callback, or directly from
libxl__initiate_device_{add,remove} if the device is already in the
desired state.
Signed-off-by: Roger Pau Monne <roger.pau@citrix.com> Acked-by: Ian Jackson <ian.jackson@eu.citrix.com> Committed-by: Ian Campbell <ian.campbell@citrix.com>
libxl: set correct nic type depending on the guest
Fix the use of nic type, which results in the following for each type
of domain:
* HVM: let the user choose, if none specified use VIF_IOEMU.
* PV: use VIF is none provided, return error if VIF_IOEMU requested.
Signed-off-by: Roger Pau Monne <roger.pau@citrix.com> Acked-by: Ian Jackson <ian.jackson@eu.citrix.com> Committed-by: Ian Campbell <ian.campbell@citrix.com>
libxl: add option to choose who executes hotplug scripts
Add and option to xl.conf file to decide if hotplug scripts are
executed from the toolstack (xl) or from udev as it used to be in the
past.
This option is only introduced in this patch, but it has no effect
since the code to call hotplug scripts from libxl is introduced in a
latter patch.
This choice will be saved in "libxl/disable_udev", as specified in the
DISABLE_UDEV_PATH constant.
Signed-off-by: Roger Pau Monne <roger.pau@citrix.com> Acked-by: Ian Jackson <ian.jackson@eu.citrix.com> Committed-by: Ian Campbell <ian.campbell@citrix.com>
libxl: convert libxl_device_nic_add to an async operation
This patch converts libxl_device_nic_add to an ao operation that
waits for device backend to reach state XenbusStateInitWait and then
marks the operation as completed. This is not really useful now, but
will be used by latter patches that will launch hotplug scripts after
we reached the desired xenbus state.
Calls to libxl_device_nic_add have also been moved to occur after the
device model has been launched, so when hotplug scripts are called
from this functions the interfaces already exists.
As usual, libxl_device_nic_add callers have been modified, and the
internal function libxl__device_disk_add has been used if the call was
inside an already running ao.
Signed-off-by: Roger Pau Monne <roger.pau@citrix.com> Acked-by: Ian Jackson <ian.jackson@eu.citrix.com> Committed-by: Ian Campbell <ian.campbell@citrix.com>
libxl: convert libxl_device_disk_add to an async op
This patch converts libxl_device_disk_add to an ao operation that
waits for device backend to reach state XenbusStateInitWait and then
marks the operation as completed. This is not really useful now, but
will be used by later patches that will launch hotplug scripts after
we reached the desired xenbus state.
As usual, libxl_device_disk_add callers have been modified, and the
internal function libxl__device_disk_add has been used if the call was
inside an already running ao.
Signed-off-by: Roger Pau Monne <roger.pau@citrix.com> Acked-by: Ian Jackson <ian.jackson@eu.citrix.com>
[ ijc -- drop hunk modifying libxl_cdrom_insert which is not needed after
25670:3666e9712eaf "libxl: make libxl_cdrom_insert async" ] Committed-by: Ian Campbell <ian.campbell@citrix.com>
libxl: convert libxl__device_disk_local_attach to an async op
This will be needed in future patches, when libxl__device_disk_add
becomes async also. Create a new status structure that defines the
local attach of a disk device and use it in
libxl__device_disk_local_attach.
This is done in this patch to split the changes introduced when
libxl__device_disk_add becomes async.
Signed-off-by: Roger Pau Monne <roger.pau@citrix.com> Acked-by: Ian Jackson <ian.jackson@eu.citrix.com> Committed-by: Ian Campbell <ian.campbell@citrix.com>
Change libxl__device_disk_add to no longer take a xs transaction and
instead pass a helper for the local attach case that's used to get the
free vdev.
This function contains some non-functional changes due to an
indentation change.
Signed-off-by: Roger Pau Monne <roger.pau@citrix.com> Acked-by: Ian Jackson <ian.jackson@eu.citrix.com> Committed-by: Ian Campbell <ian.campbell@citrix.com>
Secondary consoles are processed by libxl with the rest of the
devices by calling libxl__initiate_device_remove that waits for the
device to reach state 6 before procceeding with the removal.
When libxl is destroying the console devices, Qemu is already dead or
dying, and xenconsoled completely ignores the state backend entry for
console devices, since it performs the cleanup based on the result of
reads/writes to the tty.
Since we don't want to execute hotplug scripts for consoles, leave the
behaviour as it was previously, and just nuke the backend/frontend
xenstore entries by calling libxl__device_destroy.
Reported-by: Ian Campbell <ian.campbell@eu.citrix.com> Signed-off-by: Roger Pau Monne <roger.pau@citrix.com> Acked-by: Ian Jackson <ian.jackson@eu.citrix.com> Committed-by: Ian Campbell <ian.campbell@citrix.com>
Ian Campbell [Thu, 26 Jul 2012 14:58:18 +0000 (15:58 +0100)]
libxl: make libxl_cdrom_insert async.
This functionality is a bit of a mess and several configurations are
not properly supported.
The protocol for changing is basically to change the params node in
the disk xenstore backend. There is no interlock or error reporting in
this protocol. Completely removing the device and recreating it is not
necessary nor expected. For reference the equivalent xend code is
tools/python/xen/xend/server/blkif.py::BlkifController::reconfigureDevice().
Device model stub domains are not supported. There appears to be no
way correctly to do a media change on the emulated device while also
changing the stub domains PV backend to point to the new
backend. Reworking this is a significant task deferred until 4.3. xend
(via the equivalent "xm block-configure" functionality) also does not
support media change for stub domains (confirmed by code inspection
and experiment). Unlike xend this version errors out instead of
silently not achieving anything in this case.
There is no support for qemu-xen (upstream) media change. I expect
this is supported on the qemu side and required QMP plumbing on the
libxl side. Again this is deferred until 4.3.
On the plus side the current implementation is trivially "asynchronous".
Adds a libxl__xs_writev_atonce helper to write a key-value list to
xenstore in one go.
Tested with Windows 7.
Signed-off-by: Ian Campbell <ian.campbell@citrix.com> Acked-by: Ian Jackson <ian.jackson@eu.citrix.com> Committed-by: Ian Campbell <ian.campbell@citrix.com>
In such a way that only the cpus belonging to the cpupool of the
domain being placed are considered for the placement itself.
This happens by filtering out all the nodes in which the cpupool
has not any cpu from the placement candidates. After that ---as
cpu pooling not necessarily happens at NUMA nodes boundaries--- we
also make sure only the actual cpus that are part of the pool are
considered when counting how much processors a placement candidate
provides.
Signed-off-by: Dario Faggioli <dario.faggioli@citrix.com> Acked-by: Ian Campbell <ian.campbell@citrix.com> Acked-by: Ian Jackson <ian.jackson@eu.citrix.com> Committed-by: Ian Campbell <ian.campbell@citrix.com>
libxl: enable automatic placement of guests on NUMA nodes
If a domain does not have a VCPU affinity, try to pin it automatically
to some PCPUs. This is done taking into account the NUMA characteristics
of the host. In fact, we look for a combination of host's NUMA nodes
with enough free memory and number of PCPUs for the new domain, and pin
it to the VCPUs of those nodes.
Deciding which placement is the best happens by means of some heuristics.
For instance, smaller candidates are better, both from a domain perspective
(less memory spreading among nodes) and from the entire system perspective
(smaller memory fragmentation). In case of candidates of equal sizes
(i.e., with the same number of nodes), the amount of free memory and
the number of domains' vCPUs already pinned to the candidates' nodes are
both considered. Very often, candidates with greater amount of memory
are the one we wants, as this is good for keeping memory fragmentation
under control. However, we do not want to overcommit some node too much,
just because it has a lot of memory, and that's why the number of vCPUs
must be accounted for.
This all happens internally to libxl, and no API for driving the
mechanism is provided for now. This matches what xend already does.
Signed-off-by: Dario Faggioli <dario.faggioli@citrix.com> Acked-by: George Dunlap <george.dunlap@eu.citrix.com> Acked-by: Ian Jackson <ian.jackson@eu.citrix.com> Tested-by: Andre Przywara <andre.przywara@amd.com> Committed-by: Ian Campbell <ian.campbell@citrix.com>
libxl: libxl__xs_path_cleanup don't print error if ENOENT
Signed-off-by: Roger Pau Monne <roger.pau@citrix.com> Acked-by: Ian Campbell <ian.campbell@eu.citrix.com> Committed-by: Ian Campbell <ian.campbell@citrix.com>
Andrew Cooper [Thu, 26 Jul 2012 09:35:37 +0000 (10:35 +0100)]
xen: update xensource.com to xen.org
This patch was constructed by grepping for xensource.com over the entire
repository and eyeballing which ones were sensible to update.
In addition, the xen-tools mailing list has been deprecated, so update xentop
to refer to xen-devel instead.
Signed-off-by: Andrew Cooper <andrew.cooper3@citrix.com> Acked-by: Ian Campbell <ian.campbell@citrix.com> Committed-by: Ian Campbell <ian.campbell@citrix.com>
Ian Jackson [Thu, 26 Jul 2012 09:35:36 +0000 (10:35 +0100)]
xl: support xend empty cdrom device syntax
xend accepts `,hdc:cdrom,r' as an empty CDROM drive. However this is
not consistent with the existing xl syntax in
docs/misc/xl-disk-configuration.txt which requires `,,hdc:cdrom,r'
(the additional positional paramter is the format).
We fix this by spotting the case specially: when the target is empty
and the format contains a colon, reinterpret the format as
<vdev>:<devtype>.
Signed-off-by: Ian Campbell <ian.campbell@citrix.com> Signed-off-by: Ian Jackson <ian.jackson@eu.citrix.com> Committed-by: Ian Campbell <ian.campbell@citrix.com>
Ian Jackson [Thu, 26 Jul 2012 09:35:36 +0000 (10:35 +0100)]
xl: disk parsing preparation for empty cdrom devices
Prepare the ground for parsing the xend empty cdrom syntax, by
separating out some non-functional changes as this pre-patch:
* Clarify the disk syntax documentation wording to refer to deprecated
syntaxes too.
* Make DPC in libxlu_disk_l.l useable in the helper functions as well
as in lexer rules, by providing two definitions, each in force in
the appropriate parts of the file.
* Break the <vdev>[:<devtype>] parsing out into a helper function,
`vdev_and_devtype'.
No functional change.
Signed-off-by: Ian Jackson <ian.jackson@eu.citrix.com> Acked-by: Ian Campbell <ian.campbell@citrix.com> Committed-by: Ian Campbell <ian.campbell@citrix.com>
Ian Campbell [Thu, 26 Jul 2012 09:35:35 +0000 (10:35 +0100)]
xl: support empty CDROM devices
The important change here is to xlu_disk_parse to correctly set format == EMPTY
for CDROM devices which are empty. Test cases are added which check for
correctness here.
xend accepts ',hdc:cdrom,r'[0] as an empty CDROM drive however this is not
consistent with the xl syntax in docs/misc/xl-disk-configuration.txt which
requires ',,hdc:cdrom,r' (the additional positional paramter is the format).
I'm not sure if/how this can be fixed. Note that xend does not accept
',,hdc:cdrom,r'
There are several incidental cleanups included the the cdrom-{insert,eject}
commands:
- add a dry-run mode
- use the non-deprecated disk specification syntax
- check for and report errors from libxl_cdrom_insert
Signed-off-by: Ian Campbell <ian.campbell@citrix.com> Acked-by: Ian Jackson <ian.jackson@eu.citrix.com> Signed-off-by: Ian Jackson <ian.jackson@eu.citrix.com> Committed-by: Ian Campbell <ian.campbell@citrix.com>
projects / xen.git / log