Introduce a function to map a range of physical memory into Xen virtual
memory.
It doesn't need domheap to be setup.
It is going to be used to map the videoram.
Signed-off-by: Stefano Stabellini <stefano.stabellini@eu.citrix.com> Acked-by: Ian Campbell <ian.campbell@citrix.com> Committed-by: Ian Campbell <ian.campbell@citrix.com>
Add flush_xen_data_tlb_range_va, that flushes a range of virtual addresses.
Replace all the calls to flush_xen_data_tlb_va with calls to
flush_xen_data_tlb_range_va and remove flush_xen_data_tlb_va.
Signed-off-by: Stefano Stabellini <stefano.stabellini@eu.citrix.com> Acked-by: Ian Campbell <ian.campbell@citrix.com> Committed-by: Ian Campbell <ian.campbell@citrix.com>
Jan Beulich [Wed, 23 Jan 2013 13:19:13 +0000 (14:19 +0100)]
x86/HVM: fix RTC hour conversions
Properly mask off bit 7 when retrieving the hour values in
alarm_timer_update(), and properly use RTC_HOURS_ALARM's bit 7 when
converting from 12- to 24-hour value.
Signed-off-by: Jan Beulich <jbeulich@suse.com> Acked-by: Keir Fraser <keir@xen.org>
Jan Beulich [Wed, 23 Jan 2013 13:15:16 +0000 (14:15 +0100)]
x86: support up to 16Tb
This mainly involves adjusting the number of L4 entries needing copying
between page tables (which is now different between PV and HVM/idle
domains), and changing the cutoff point and method when more than the
supported amount of memory is found in a system.
Since TMEM doesn't currently cope with the full 1:1 map not always
being visible, it gets forcefully disabled in that case.
Signed-off-by: Jan Beulich <jbeulich@suse.com> Acked-by: Keir Fraser <keir@xen.org> Acked-by: Dan Magenheimer <dan.magenheimer@oracle.com>
Jan Beulich [Wed, 23 Jan 2013 13:14:34 +0000 (14:14 +0100)]
tmem: partial adjustments for x86 16Tb support
Despite the changes below, tmem still has code assuming to be able to
directly access all memory, or mapping arbitrary amounts of not
directly accessible memory. I cannot see how to fix this without
converting _all_ its domheap allocations to xenheap ones. And even then
I wouldn't be certain about there not being other cases where the "all
memory is always mapped" assumption would be broken. Therefore, tmem
gets disabled by the next patch for the time being if the full 1:1
mapping isn't always visible.
Signed-off-by: Jan Beulich <jbeulich@suse.com> Acked-by: Dan Magenheimer <dan.magenheimer@oracle.com> Acked-by: Keir Fraser <keir@xen.org>
Jan Beulich [Wed, 23 Jan 2013 13:12:54 +0000 (14:12 +0100)]
x86: properly use map_domain_page() in nested HVM code
This eliminates a couple of incorrect/inconsistent uses of
map_domain_page() from VT-x code.
Note that this does _not_ add error handling where none was present
before, even though I think NULL returns from any of the mapping
operations touched here need to properly be handled. I just don't know
this code well enough to figure out what the right action in each case
would be.
Signed-off-by: Jan Beulich <jbeulich@suse.com> Acked-by: Keir Fraser <keir@xen.org>
Jan Beulich [Wed, 23 Jan 2013 13:08:44 +0000 (14:08 +0100)]
x86: properly use map_domain_page() when building Dom0
This requires a minor hack to allow the correct page tables to be used
while running on Dom0's page tables (as they can't be determined from
"current" at that time).
Signed-off-by: Jan Beulich <jbeulich@suse.com> Acked-by: Keir Fraser <keir@xen.org>
Jan Beulich [Wed, 23 Jan 2013 13:06:20 +0000 (14:06 +0100)]
x86: re-introduce map_domain_page() et al
This is being done mostly in the form previously used on x86-32,
utilizing the second L3 page table slot within the per-domain mapping
area for those mappings. It remains to be determined whether that
concept is really suitable, or whether instead re-implementing at least
the non-global variant from scratch would be better.
Also add the helpers {clear,copy}_domain_page() as well as initial uses
of them.
One question is whether, to exercise the non-trivial code paths, we
shouldn't make the trivial shortcuts conditional upon NDEBUG being
defined. See the debugging patch at the end of the series.
Signed-off-by: Jan Beulich <jbeulich@suse.com> Acked-by: Keir Fraser <keir@xen.org>
Fix S3 regression introduced by cs 23013:65d26504e843 (ACPI: large
cleanup). The dmar virtual pointer returned from acpi_get_table cannot
be safely stored away and used later, as the underlying
acpi_os_map_memory / __acpi_map_table functions overwrite the mapping
causing it to point to different tables than dmar (last fetched table is
used). This subsequently causes acpi_dmar_reinstate() and
acpi_dmar_zap() to write data to wrong table, causing its corruption and
problems with consecutive s3 resumes.
Added a new function to fetch ACPI table physical address, and
establishing separate static mapping for dmar_table pointer instead of
using acpi_get_table().
Signed-off-by: Tomasz Wroblewski <tomasz.wroblewski@citrix.com>
Added call to acpi_tb_verify_table(). Fixed page count passed to
map_pages_to_xen(). Cosmetic changes.
Signed-off-by: Jan Beulich <jbeulich@suse.com> Committed-by: Jan Beulich <jbeulich@suse.com>
Daniel De Graaf [Wed, 23 Jan 2013 09:18:50 +0000 (09:18 +0000)]
xen/arch/x86: complete XSM hooks on irq/pirq mappings
Manipulation of a domain's pirq namespace was not fully protected by
XSM hooks because the XSM hooks for IRQs needed a physical IRQ. Since
this may not apply to HVM domains, a complete solution needs to split
the XSM hook for this operation, using one hook for the PIRQ
manipulation and one for controlling access to the hardware IRQ.
This reworking has the advantage of providing the same MSI data to
remove_irq that is provided to add_irq, allowing the PCI device to be
determined in both functions. It also eliminates the last callers of
rcu_lock_target_domain_by_id in x86 and common code in preparation for
this function's removal.
Signed-off-by: Daniel De Graaf <dgdegra@tycho.nsa.gov> Committed-by: Keir Fraser <keir@xen.org>
Daniel De Graaf [Wed, 23 Jan 2013 09:17:19 +0000 (09:17 +0000)]
hvm: wire up domctl and xsm hypercalls
These hypercalls are usable by HVM guests. Once connected, simple
functions of the Xen toolstack can be run from an HVM domain if that
domain is permitted access (which is currently only possible via XSM).
Signed-off-by: Daniel De Graaf <dgdegra@tycho.nsa.gov> Committed-by: Keir Fraser <keir@xen.org>
Jan Beulich [Tue, 22 Jan 2013 08:33:10 +0000 (09:33 +0100)]
x86: restore (optional) forwarding of PCI SERR induced NMI to Dom0
c/s 22949:54fe1011f86b removed the forwarding of NMIs to Dom0 when they
were caused by PCI SERR. NMI buttons as well as BMCs (like HP's iLO)
may however want such events to be seen in Dom0 (e.g. to trigger a
dump).
Therefore restore most of the functionality which named c/s removed
(adjusted for subsequent changes, and adjusting the public interface to
use the modern term, retaining the old one for backwards
compatibility).
Ian Campbell [Mon, 21 Jan 2013 16:04:56 +0000 (16:04 +0000)]
vtpmmgr: fix build on 32-bit
Correct format string, fixing:
vtpm_storage.c: In function 'vtpm_storage_load_header': vtpm_storage.c:658: error: format '%ld' expects type 'long int', but argument 5 has type 'unsigned int'
vtpm_storage.c:658: error: format '%ld' expects type 'long int', but argument 5 has type 'unsigned int' make[2]: *** [vtpm_storage.o] Error 1
Add padlock.o to PSSL_OBJS, fixing:
/local/scratch/ianc/devel/xen-unstable.git/stubdom/mini-os-x86_32-vtpmmgr/mini-os.o: In function `aes_crypt_ecb': /local/scratch/ianc/devel/xen-unstable.git/stubdom/polarssl-x86_32/library/aes.c:659: undefined reference to `padlock_supports'
/local/scratch/ianc/devel/xen-unstable.git/stubdom/polarssl-x86_32/library/aes.c:661: undefined reference to `padlock_xcryptecb' /local/scratch/ianc/devel/xen-unstable.git/stubdom/mini-os-x86_32-vtpmmgr/mini-os.o: In function `aes_crypt_cbc': /local/scratch/ianc/devel/xen-unstable.git/stubdom/polarssl-x86_32/library/aes.c:771: undefined reference to `padlock_supports'
/local/scratch/ianc/devel/xen-unstable.git/stubdom/polarssl-x86_32/library/aes.c:773: undefined reference to `padlock_xcryptcbc'
make[1]: ***
[/local/scratch/ianc/devel/xen-unstable.git/stubdom/mini-os-x86_32-vtpmmgr/mini-os]
Error 1
Signed-off-by: Ian Campbell <ian.campbell@citrix.com>
Acked by: Matthew Fioravante <matthew.fioravante@jhuapl.edu>
[ ijc -- applied same fix to stubdom/vtpm/Makefile ] Committed-by: Ian Campbell <ian.campbell@citrix.com>
xen/arm: flush dcache after memcpy'ing the kernel image
After memcpy'ing the kernel in guest memory we need to flush the dcache
to make sure that the data actually reaches the memory before we start
executing guest code with caches disabled.
copy_from_paddr is the function that does the copy, so add a
flush_xen_dcache_va_range there.
Signed-off-by: Stefano Stabellini <stefano.stabellini@eu.citrix.com> Acked-by: Ian Campbell <ian.campbell@citrix.com> Committed-by: Ian Campbell <ian.campbell@citrix.com>
Ian Campbell [Mon, 21 Jan 2013 12:40:31 +0000 (12:40 +0000)]
arm: use module provided command line for domain 0 command line
Fallback to xen,dom0-bootargs if this isn't present.
Ideally this would use module1-args iff the kernel came from the
modules and the existing xen,dom0-bootargs if the kernel came from
flash, but this approach is simpler and has the same effect in
practice.
Signed-off-by: Ian Campbell <ian.campbell@citrix.com> Acked-by: Tim Deegan <tim@xen.org> Committed-by: Ian Campbell <ian.campbell@citrix.com>
Ian Campbell [Mon, 21 Jan 2013 12:40:27 +0000 (12:40 +0000)]
arm: avoid placing Xen over any modules.
This will still fail if the modules are such that Xen is pushed out of
the top 32M of RAM since it will then overlap with the domheap (or
possibly xenheap). This will be dealt with later.
Signed-off-by: Ian Campbell <ian.campbell@citrix.com> Acked-by: Tim Deegan <tim@xen.org> Committed-by: Ian Campbell <ian.campbell@citrix.com>
Ian Campbell [Mon, 21 Jan 2013 12:40:26 +0000 (12:40 +0000)]
xen: arm: introduce concept of modules which can be in RAM at start of day
The intention is that these will eventually be filled in with
information from the bootloader, perhaps via a DTB binding.
Allow for 2 modules (kernel and initrd), plus a third pseudo-module
which is the hypervisor itself. Currently we neither parse nor do
anything with them.
Signed-off-by: Ian Campbell <ian.campbell@citrix.com> Acked-by: Stefano Stabellini <stefano.stabellini@eu.citrix.com> Committed-by: Ian Campbell <ian.campbell@citrix.com>
Tim Deegan [Fri, 18 Jan 2013 11:31:57 +0000 (12:31 +0100)]
x86/hvm: fix RTC setting.
When the guest writes one field of the RTC time, we must bring all the
other fields up to date for the current second before calculating the
new RTC time.
Signed-off-by: Tim Deegan <tim@xen.org> Tested-by: Phil Evans <Phil.Evans@m247.com> Committed-by: Jan Beulich <jbeulich@suse.com>
Boris Ostrovsky [Fri, 18 Jan 2013 11:20:58 +0000 (12:20 +0100)]
x86/AMD: Enable WC+ memory type on family 10 processors
In some cases BIOS may not enable WC+ memory type on family 10 processors,
instead converting what would be WC+ memory to CD type. On guests using
nested pages this could result in performance degradation. This patch
enables WC+.
Signed-off-by: Boris Ostrovsky <boris.ostrovsky@amd.com> Committed-by: Jan Beulich <jbeulich@suse.com>
Add conditional build of subsystems to configure.ac
The toplevel Makefile still works without running configure
and will default build everything
Signed-off-by: Matthew Fioravante <matthew.fioravante@jhuapl.edu> Acked-by: Ian Campbell <ian.campbell@citrix.com> Committed-by: Ian Campbell <ian.campbell@citrix.com>
Signed-off-by: Matthew Fioravante <matthew.fioravante@jhuapl.edu> Acked-by: Ian Campbell <ian.campbell@citrix.com> Committed-by: Ian Campbell <ian.campbell@citrix.com>
Stub domains now use autoconf to build.
This configure script can enable or disable specific domains
and also specify custom download locations for stubdom library
packages. See ./configure --help for details.
C and Caml are disabled by default. vtpm-stubdom is conditional
on the presense of cmake.
Rename vtpmmgrdom to vtpmmgr-stubdom
Also update .*ignore
Signed-off-by: Matthew Fioravante <matthew.fioravante@jhuapl.edu> Acked-by: Ian Campbell <ian.campbell@citrix.com> Committed-by: Ian Campbell <ian.campbell@citrix.com>
Signed-off-by: Matthew Fioravante <matthew.fioravante@jhuapl.edu> Acked-by: Ian Campbell <ian.campbell@citrix.com> Committed-by: Ian Campbell <ian.campbell@citrix.com>
Signed-off-by: Matthew Fioravante <matthew.fioravante@jhuapl.edu> Acked-by: Ian Campbell <ian.campbell@citrix.com> Committed-by: Ian Campbell <ian.campbell@citrix.com>
vtpm/vtpmmgr and required libs to stubdom/Makefile
Add 3 new libraries to stubdom:
libgmp
polarssl
Berlios TPM Emulator 0.7.4
Add makefile structure for vtpm and vtpmmgrdom. Both
vtpm domains are optional builds as vtpm depends on
cmake. To build either of them, you must do so explicitly.
make vtpm-stubdom vtpmmgrdom
Finally, also update .*ignore
Signed-off-by: Matthew Fioravante <matthew.fioravante@jhuapl.edu> Acked-by: Ian Campbell <ian.campbell@citrix.com>
[ijc, folded in fix from Matthew to workaround cmake 2.8.2 build failure] Committed-by: Ian Campbell <ian.campbell@citrix.com>
Add the code base for vtpmmgrdom. Makefile changes
next patch.
Signed-off-by: Matthew Fioravante <matthew.fioravante@jhuapl.edu> Acked-by: Ian Campbell <ian.campbell@citrix.com> Committed-by: Ian Campbell <ian.campbell@citrix.com>
Add the code base for vtpm-stubdom to the stubdom
heirarchy. Makefile changes in later patch.
Signed-off-by: Matthew Fioravante <matthew.fioravante@jhuapl.edu> Acked-by: Ian Campbell <ian.campbell@citrix.com> Committed-by: Ian Campbell <ian.campbell@citrix.com>
Paolo Bonzini [Fri, 18 Jan 2013 10:35:11 +0000 (11:35 +0100)]
x86: find a better location for the real-mode trampoline
On some machines, the location at 0x40e does not point to the beginning
of the EBDA. Rather, it points to the beginning of the BIOS-reserved
area of the EBDA, while the option ROMs place their data below that
segment.
For this reason, 0x413 is actually a better source than 0x40e to get
the location of the real-mode trampoline. Xen was already using it
as a second source, and this patch keeps that working. However, just
in case, let's also fetch the information from the multiboot structure,
where the boot loader should have placed it. This way we don't
necessarily trust one of the BIOS or the multiboot loader more than
the other.
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
Retain the previous code, thus using the multiboot value only if it's
sane but lower than the BDA computed one. Also use the full 32-bit
mem_lower value and prefer MBI_MEMLIMITS over open coding it (requiring
a slight adjustment to multiboot.h to make its constants actually
usable in assembly code, which previously they were only meant to be).
Signed-off-by: Jan Beulich <jbeulich@suse.com> Acked-by: Keir Fraser <keir@xen.org> Committed-by: Jan Beulich <jbeulich@suse.com>
xen/arm: initialize the GIC irq properties of interrupts routed to guests
We are currently initializing GIC irq properties (ITARGETSR, IPRIORITYR,
and GICD_ICFGR) only in gic_route_irq, that is not called for guest
interrupts at all.
Move the initialization into a separate function
(gic_set_irq_properties) and call it from gic_route_irq_to_guest.
Signed-off-by: Stefano Stabellini <stefano.stabellini@eu.citrix.com> Acked-by: Ian Campbell <ian.campbell@citrix.com> Committed-by: Ian Campbell <ian.campbell@citrix.com>
Save and restore the virtual timer registers during the context switch.
At save time initialize an internal Xen timer to make sure that Xen
schedules the guest vcpu at the time of the next virtual timer
interrupt.
Receive the virtual timer interrupt into the hypervisor and inject it
into the running guest.
Signed-off-by: Stefano Stabellini <stefano.stabellini@eu.citrix.com> Acked-by: Ian Campbell <ian.campbell@citrix.com> Committed-by: Ian Campbell <ian.campbell@citrix.com>
Ian Campbell [Thu, 17 Jan 2013 16:48:23 +0000 (16:48 +0000)]
xen: return a per-mapping error from XENMEM_add_to_physmap_range.
Since ARM and PVH dom0 kernel use this to map foreign domain pages
they could in the future hit paged out or shared pages etc and
therefore need to propagate which frames are -ENOENT and which failed
for some other reason.
We have not yet released a version of Xen with this particular
hypercall subop so we can change the interface without worrying about
compatibility (I think/hope).
This would be used by the privcmd driver, in particular it relates to
Mats' patch "improve performance of MMAPBATCH_V2."
NB I have only implemented the ARM side since the PVH side isn't in
tree yet.
Signed-off-by: Ian Campbell <ian.campbell@citrix.com> Acked-by: Stefano Stabellini <stefano.stabellini@eu.citrix.com> Acked-by: Keir Fraser <keir@xen.org> Acked-by: Mats Petersson <mats.petersson@citrix.com> Committed-by: Ian Campbell <ian.campbell@citrix.com>
Lars Rasmusson [Thu, 17 Jan 2013 16:48:22 +0000 (16:48 +0000)]
xen: arm: Correct register values and comment in early init_uart.
Set register values and comment in early init_uart to match
documentation of PL011 UART
Reading the PL011 UART documentation on
http://infocenter.arm.com/help/topic/com.arm.doc.ddi0183f/DDI0183.pdf
in sec 3.2 shows the early initialisation of the UART on the Versatile Express
is incorrect. This fixes it.
Signed-off-by: Lars Rasmusson <Lars.Rasmusson@sics.se> Acked-by: Ian Campbell <ian.campbell@citrix.com> Committed-by: Ian Campbell <ian.campbell@citrix.com>
Daniel De Graaf [Thu, 17 Jan 2013 16:48:21 +0000 (16:48 +0000)]
libxl: correct xenstore permissions on console device
When the console is connected to a domain other than dom0, the console
device's backend field should be set so the xenstore permissions for the
console device reflect the domain that will be accessing them.
Signed-off-by: Daniel De Graaf <dgdegra@tycho.nsa.gov> Acked-by: Ian Campbell <ian.campbell@citrix.com> Committed-by: Ian Campbell <ian.campbell@citrix.com>
Daniel De Graaf [Thu, 17 Jan 2013 16:48:21 +0000 (16:48 +0000)]
xenconsoled: use grant references instead of map_foreign_range
Grant references for the xenstore and xenconsole shared pages exist, but
currently only xenstore uses these references. Change the xenconsole
daemon to prefer using the grant reference over map_foreign_range when
mapping the shared console ring.
This allows xenconsoled to be run in a domain other than dom0 if set up
correctly - for libxl, the xenstore path /tool/xenconsoled/domid
specifies the domain containing xenconsoled.
Signed-off-by: Daniel De Graaf <dgdegra@tycho.nsa.gov> Acked-by: Ian Campbell <ian.campbell@citrix.com> Committed-by: Ian Campbell <ian.campbell@citrix.com>
Ian Campbell [Thu, 17 Jan 2013 13:53:14 +0000 (13:53 +0000)]
tools: Update to SeaBIOS 1.7.1
Only lightly tested with a Linux HVM guest PXE boot.
Accept the defaults for the config options. Many of them are not
relevant to Xen but this matches what others (at least the Debian
SeaBIOS packages and the binary shipped by Qemu) are doing. The
Debian Xen packages are built against Debian's SeaBIOS package so
there is value in being similar.
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>
Julien Grall [Thu, 17 Jan 2013 13:53:09 +0000 (13:53 +0000)]
libxl: don't continue to create the domain if the device model is not spawned
When the device model can't be spawned, rc variable is cleared in
device_model_spawn_outcome (libxl_dm.c).
In this case libxl will continue to create the domain and let it between life
and death.
Signed-off-by: Julien Grall <julien.grall@citrix.com> Acked-by: Ian Campbell <ian.campbell@citrix.com> Committed-by: Ian Campbell <ian.campbell@citrix.com>
If we pass 0 as pygrub --entry argument (i.e. we want to boot first item), default value is used instead. This is dueto wrong check for range of allowed values of index - 0 is index of first item.
Signed-off-by: Miroslav Rezanina <mrezanin@redhat.com> Acked-by: Matt Wilson <msw@amazon.com> Acked-by: Ian Campbell <ian.campbell@citrix.com> Committed-by: Ian Campbell <ian.campbell@citrix.com>
Wei Liu [Thu, 17 Jan 2013 13:53:03 +0000 (13:53 +0000)]
Switch from select() to poll() in xenconsoled's IO loop
In Linux select() typically supports up to 1024 file descriptors. This can be
a problem when user tries to boot up many guests. Switching to poll() has
minimum impact on existing code and has better scalibility.
pollfd array is dynamically allocated / reallocated. If the array fails to
expand, we just ignore the incoming fd.
Updated: reset *_pollfd after use.
This fixes regression 14869.
Also remove unused slave_pollfd in strcut domain.
Signed-off-by: Wei Liu <wei.liu2@citrix.com> Acked-by: Ian Campbell <ian.campbell@citrix.com> Committed-by: Ian Campbell <ian.campbell@citrix.com>
Razvan Cojocaru [Thu, 17 Jan 2013 12:27:00 +0000 (12:27 +0000)]
mem_event: Add support for MEM_EVENT_REASON_MSR
Add the new MEM_EVENT_REASON_MSR event type. Works similarly
to the other register events, except event.gla always contains
the MSR address (in addition to event.gfn, which holds the value).
MEM_EVENT_REASON_MSR does not honour the HVMPME_onchangeonly bit,
as doing so would complicate the hvm_msr_write_intercept()
switch-based handling of writes for different MSR addresses,
with little added benefit.
Signed-off-by: Razvan Cojocaru <rzvncj@gmail.com> Acked-by: Tim Deegan <tim@xen.org> Committed-by: Tim Deegan <tim@xen.org>
Robert Phillips [Thu, 17 Jan 2013 11:53:42 +0000 (11:53 +0000)]
x86/mm: Provide support for multiple frame buffers in HVM guests.
Support is provided for both shadow and hardware assisted paging (HAP)
modes. This code bookkeeps the set of video frame buffers (vram),
detects when the guest has modified any of those buffers and, upon request,
returns a bitmap of the modified pages.
This lets other software components re-paint the portions of the monitor
(or monitors) that have changed.
Each monitor has a frame buffer of some size at some position
in guest physical memory.
The set of frame buffers being tracked can change over time as monitors
are plugged and unplugged.
Signed-off-by: Robert Phillips <robert.phillips@citrix.com> Acked-by: Tim Deegan <tim@xen.org>
Removed a stray #include and a few hard tabs.
Signed-off-by: Tim Deegan <tim@xen.org> Committed-by: Tim Deegan <tim@xen.org>
Jan Beulich [Thu, 17 Jan 2013 09:56:34 +0000 (10:56 +0100)]
miscellaneous cleanup
... noticed while putting together the 16Tb support patches for x86.
Briefly, this (in order of the changes below)
- fixes an inefficiency in x86's context switch code (translations to/
from struct page are more involved than to/from MFNs)
- drop unnecessary MFM-to-page conversions
- drop a redundant call to destroy_xen_mappings() (an indentical call
is being made a few lines up)
- simplify a VA-to-MFN translation
- drop dead code (several occurrences)
- add a missing __init annotation
Signed-off-by: Jan Beulich <jbeulich@suse.com> Acked-by: Keir Fraser <keir@xen.org>
Jan Beulich [Thu, 17 Jan 2013 09:55:00 +0000 (10:55 +0100)]
x86/EFI: retrieve PCI ROM contents not accessible through BARs
Linux 3.8-rc added code to do this, so we need to support this in the
hypervisor for Dom0 to be able to get at the same information as a
native kernel.
Signed-off-by: Jan Beulich <jbeulich@suse.com> Acked-by: Keir Fraser <keir@xen.org>
Jan Beulich [Wed, 16 Jan 2013 12:56:55 +0000 (13:56 +0100)]
x86: consistently mask floating point exceptions
c/s 23142:f5e8d152a565 resulted in v->arch.fpu_ctxt to point into the
save area allocated for xsave/xrstor (when they're available). The way
vcpu_restore_fpu_lazy() works (using fpu_init() for an uninitialized
vCPU only when there's no xsave support) causes this to load whatever
arch_set_info_guest() put there, irrespective of whether the i387 state
was specified to be valid in the respective input structure.
Consequently, with a cleared (al zeroes) incoming FPU context, and with
xsave available, one gets all exceptions unmasked (as opposed to to the
legacy case, where FINIT and LDMXCSR get used, masking all exceptions).
This causes e.g. para-virtualized NetWare to crash.
The behavior of arch_set_info_guest() is thus being made more hardware-
like for the FPU portion of it: Considering it to be similar to INIT,
it will leave untouched all floating point state now. An alternative
would be to make the behavior RESET-like, forcing all state to known
values, albeit - taking into account legacy behavior - not to precisely
the values RESET would enforce (which masks only SSE exceptions, but
not x87 ones); that would come closest to mimicing FINIT behavior in
the xsave case. Another option would be to continue copying whatever
was provided, but override (at least) FCW and MXCSR if VGCF_I387_VALID
isn't set.
Signed-off-by: Jan Beulich <jbeulich@suse.com> Acked-by: Keir Fraser <keir@xen.org>
Zhang Xiantao [Tue, 15 Jan 2013 10:33:41 +0000 (11:33 +0100)]
nEPT: Expose EPT & VPID capablities to L1 VMM
Expose EPT's and VPID 's basic features to L1 VMM.
For EPT, no EPT A/D bit feature supported.
For VPID, exposes all features to L1 VMM
Signed-off-by: Zhang Xiantao <xiantao.zhang@intel.com> Acked-by: Tim Deegan <tim@xen.org> Acked-by: Jun Nakajima <jun.nakajima@intel.com> Acked-by: Eddie Dong <eddie.dong@intel.com> Committed-by: Jan Beulich <jbeulich@suse.com>
Zhang Xiantao [Tue, 15 Jan 2013 10:30:50 +0000 (11:30 +0100)]
nVMX: virutalize VPID capability to nested VMM
Virtualize VPID for the nested vmm, use host's VPID
to emualte guest's VPID. For each virtual vmentry, if
guest'v vpid is changed, allocate a new host VPID for
L2 guest.
Signed-off-by: Zhang Xiantao <xiantao.zhang@intel.com> Acked-by: Tim Deegan <tim@xen.org> Acked-by: Jun Nakajima <jun.nakajima@intel.com> Acked-by: Eddie Dong <eddie.dong@intel.com> Committed-by: Jan Beulich <jbeulich@suse.com>
Zhang Xiantao [Tue, 15 Jan 2013 10:29:41 +0000 (11:29 +0100)]
nEPT: handle invept instruction from L1 VMM
Add the INVEPT instruction emulation logic.
Signed-off-by: Zhang Xiantao <xiantao.zhang@intel.com> Acked-by: Tim Deegan <tim@xen.org> Acked-by: Jun Nakajima <jun.nakajima@intel.com> Acked-by: Eddie Dong <eddie.dong@intel.com> Committed-by: Jan Beulich <jbeulich@suse.com>
Zhang Xiantao [Tue, 15 Jan 2013 10:28:23 +0000 (11:28 +0100)]
nEPT: Use minimal permission for nested p2m
Emulate permission check for the nested p2m. Current solution is to
use minimal permission, and once meet permission violation in L0, then
determin whether it is caused by guest EPT or host EPT
Signed-off-by: Zhang Xiantao <xiantao.zhang@intel.com> Acked-by: Tim Deegan <tim@xen.org> Acked-by: Jun Nakajima <jun.nakajima@intel.com> Acked-by: Eddie Dong <eddie.dong@intel.com> Committed-by: Jan Beulich <jbeulich@suse.com>
Zhang Xiantao [Tue, 15 Jan 2013 10:23:05 +0000 (11:23 +0100)]
nEPT: Sync PDPTR fields if L2 guest in PAE paging mode
For PAE L2 guest, GUEST_DPPTR registers needs to be synced for each virtual
vmentry.
Signed-off-by: Zhang Xiantao <xiantao.zhang@intel.com> Acked-by: Tim Deegan <tim@xen.org> Acked-by: Jun Nakajima <jun.nakajima@intel.com> Acked-by: Eddie Dong <eddie.dong@intel.com> Committed-by: Jan Beulich <jbeulich@suse.com>
Zhang Xiantao [Tue, 15 Jan 2013 10:18:46 +0000 (11:18 +0100)]
nEPT: Try to enable EPT paging for L2 guest
Once found EPT is enabled by L1 VMM, enabled nested EPT support
for L2 guest.
Signed-off-by: Zhang Xiantao <xiantao.zhang@intel.com> Acked-by: Tim Deegan <tim@xen.org> Acked-by: Jun Nakajima <jun.nakajima@intel.com> Acked-by: Eddie Dong <eddie.dong@intel.com> Committed-by: Jan Beulich <jbeulich@suse.com>
Zhang Xiantao [Tue, 15 Jan 2013 10:17:01 +0000 (11:17 +0100)]
EPT: Make ept data structure or operations neutral
Share the current EPT logic with nested EPT case, so
make the related data structure or operations netural
to comment EPT and nested EPT.
Signed-off-by: Zhang Xiantao <xiantao.zhang@intel.com> Acked-by: Tim Deegan <tim@xen.org> Acked-by: Jun Nakajima <jun.nakajima@intel.com> Acked-by: Eddie Dong <eddie.dong@intel.com> Committed-by: Jan Beulich <jbeulich@suse.com>
Zhang Xiantao [Tue, 15 Jan 2013 10:15:29 +0000 (11:15 +0100)]
nested_ept: Implement guest ept's walker
Implment guest EPT PT walker, some logic is based on shadow's
ia32e PT walker. During the PT walking, if the target pages are
not in memory, use RETRY mechanism and get a chance to let the
target page back.
Signed-off-by: Zhang Xiantao <xiantao.zhang@intel.com> Acked-by: Tim Deegan <tim@xen.org> Acked-by: Jun Nakajima <jun.nakajima@intel.com> Acked-by: Eddie Dong <eddie.dong@intel.com> Committed-by: Jan Beulich <jbeulich@suse.com>
Zhang Xiantao [Tue, 15 Jan 2013 10:11:37 +0000 (11:11 +0100)]
nestedhap: Change nested p2m's walker to vendor-specific
EPT and NPT adopts differnt formats for each-level entry,
so change the walker functions to vendor-specific.
Signed-off-by: Zhang Xiantao <xiantao.zhang@intel.com> Acked-by: Tim Deegan <tim@xen.org> Acked-by: Jun Nakajima <jun.nakajima@intel.com> Acked-by: Eddie Dong <eddie.dong@intel.com> Committed-by: Jan Beulich <jbeulich@suse.com>
Zhang Xiantao [Tue, 15 Jan 2013 10:09:33 +0000 (11:09 +0100)]
nestedhap: Change hostcr3 and p2m->cr3 to meaningful words
VMX doesn't have the concept about host cr3 for nested p2m,
and only SVM has, so change it to netural words.
Signed-off-by: Zhang Xiantao <xiantao.zhang@intel.com> Acked-by: Tim Deegan <tim@xen.org> Acked-by: Jun Nakajima <jun.nakajima@intel.com> Acked-by: Eddie Dong <eddie.dong@intel.com> Committed-by: Jan Beulich <jbeulich@suse.com>
Dongxiao Xu [Fri, 11 Jan 2013 16:30:56 +0000 (17:30 +0100)]
nested vmx: fix CR0/CR4 emulation
While emulate CR0 and CR4 for nested virtualization, set the CR0/CR4
guest host mask to 0xffffffff in shadow VMCS, then calculate the
corresponding read shadow separately for CR0 and CR4. While getting
the VM exit for CR0/CR4 access, check if L1 VMM owns the bit. If so,
inject the VM exit to L1 VMM. Otherwise, L0 will handle it and sync
the value to L1 virtual VMCS.
Signed-off-by: Dongxiao Xu <dongxiao.xu@intel.com> Acked-by: Eddie Dong <eddie.dong@intel.com> Committed-by: Jan Beulich <jbeulich@suse.com>
Wei Liu [Fri, 11 Jan 2013 12:22:30 +0000 (12:22 +0000)]
Switch from select() to poll() in xenconsoled's IO loop
In Linux select() typically supports up to 1024 file descriptors. This can be
a problem when user tries to boot up many guests. Switching to poll() has
minimum impact on existing code and has better scalibility.
pollfd array is dynamically allocated / reallocated. If the array fails to
expand, we just ignore the incoming fd.
Signed-off-by: Wei Liu <wei.liu2@citrix.com> Acked-by: Ian Campbell <ian.campbell@citrix.com> Committed-by: Ian Campbell <ian.campbell@citrix.com>
Some renaming to correct the PCI and SBDF terminology.
No functional change.
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>
Andrew Cooper [Fri, 11 Jan 2013 12:22:29 +0000 (12:22 +0000)]
tools/ocaml: libxc bindings: Fix SBDF encoding
Changeset 23861:ec7c81fbe0de alters the SBDF encoding expected by the
DOMCTL_{de,}assign_device hypercalls.
While it updates libxl, libxc and the python bindings, the ocaml
bindings got missed. As a result, any attempt to use PCI Passthrough
with Xen-4.2 and later will fail.
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>
fix wrong path while calling pygrub and libxl-save-helper
in current xen x86_64, the default libexec directory is /usr/lib/xen/bin,
while the private binder is /usr/lib64/xen/bin. but some commands(pygrub,
libxl-save-helper) located in private binder directory is called from
libexec directory which lead to the following error:
1, for pygrub bootloader:
libxl: debug: libxl_bootloader.c:429:bootloader_disk_attached_cb: /usr/lib/xen/bin/pygrub doesn't exist, falling back to config path
2, for libxl-save-helper:
libxl: cannot execute /usr/lib/xen/bin/libxl-save-helper: No such file or directory
libxl: error: libxl_utils.c:363:libxl_read_exactly: file/stream truncated reading ipc msg header from domain 3 save/restore helper stdout pipe
libxl: error: libxl_exec.c:118:libxl_report_child_exitstatus: domain 3 save/restore helper [10222] exited with error status 255
there are two ways to fix above error. the first one is make such command
store in the /usr/lib/xen/bin and /usr/lib64/xen/bin(symbol link to
previous), e.g. qemu-dm. The second way is using private binder dir
instead of libexec dir. e.g. xenconsole.
For these cases, the latter one is suitable.
Signed-off-by: Bamvor Jian Zhang <bjzhang@suse.com> Committed-by: Ian Campbell <ian.campbell@citrix.com>
Ian Campbell [Fri, 11 Jan 2013 12:22:27 +0000 (12:22 +0000)]
libxc: x86: ensure that the initial mapping fits into the guest's memory
In particular we need to check that adding 512KB of slack and
rounding up to a 4MB boundary do not overflow the guest's memory
allocation. Otherwise we run off the end of the p2m when building the
guest's initial page tables and populate them with garbage.
Wei noticed this when build tiny (2MB) mini-os domains.
Reported-by: Wei Liu <Wei.Liu2@citrix.com> Signed-off-by: Ian Campbell <ian.campbell@citrix.com> Acked-by: Jan Beulich <jbeulich@suse.com> Committed-by: Ian Campbell <ian.campbell@citrix.com>
Jim Fehlig [Fri, 11 Jan 2013 12:22:26 +0000 (12:22 +0000)]
libxl: Set vfb and vkb devid if not done so by the caller
Other devices set a sensible devid if the caller has not done so.
Do the same for vfb and vkb. While at it, factor out the common code
used to determine a sensible devid, so it can be used by other
libxl__device_*_add functions.
Signed-off-by: Jim Fehlig <jfehlig@suse.com> Acked-by: Ian Campbell <ian.campbell@citrix.com> Committed-by: Ian Campbell <ian.campbell@citrix.com>
Daniel De Graaf [Fri, 11 Jan 2013 10:49:49 +0000 (10:49 +0000)]
xsm/flask: document the access vectors
This adds comments to the FLASK access_vectors file describing what
operations each access vector controls and the meanings of the source
and target fields in the permission check. This also makes the
indentation of the file consistent; no functionality changes are made.
Signed-off-by: Daniel De Graaf <dgdegra@tycho.nsa.gov> Committed-by: Keir Fraser <keir@xen.org>
projects / xen.git / log