arm: do not set max_vcpus = 8 in arch_domain_create.

XEN_DOMCTL_max_vcpus cannot reduce max_vcpus and therefore we can't create a
smaller guest.

The limit of 8 (due to GIC limits) should be expressed in MAX_VIRT_CPUS.

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>

arm: implement stub version of flush_tlb_mask.

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>

arm: stub out sync_vcpu_execstate

We don't do lazy exec state switching so there isn't actually anything to do.

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>

arm: remove hard tabs from init_idle_domain

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>

arm: implement p2m lookup

Signed-off-by: Ian Campbell <ian.campbell@citrix.com>
Acked-by: Tim Deegan <tim@xen.org>
Committed-by: Ian Campbell <ian.campbell@citrix.com>

arm: remove unnecessarily verbose print from p2m_load_VTTBR

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>

arm: print domid as part of debug trap

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>

arm: allocate and setup a guest vcpu.

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>

arm: hook up domctl and memory_op

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>

arm: enable interrupts while handling traps

For most traps we can do this as soon as we have saved the necessary state.
For IRQs and FIQs we must wait until we have acked the interrupt with the GIC.

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>

arm: restore stack on return from trap.

We align the stack before calling into C code but we weren't undoing this on
return.

Collapse continue_(non)idle_domain into continue_new_vcpu.

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>

arm: correct and expand TLB flush CP15 registers

Correct spelling of TLBIALLHIS and correct definition of TLBIALLNSNHIS.

Add a few more.

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>

arm: handy function to print a walk of a page table

Include helpers for dumping hypervisor walks and guest p2m walks.

Useful for debug but not actually used in this patch.

Signed-off-by: Ian Campbell <ian.campbell@citrix.com>
Acked-by: Tim Deegan <tim@xen.org>
Committed-by: Ian Campbell <ian.campbell@citrix.com>

arm: allocate top level p2m page for all non-idle domains

Not just dom0.

Signed-off-by: Ian Campbell <ian.campbell@citrix.com>
Acked-by: Tim Deegan <tim@xen.org>
Committed-by: Ian Campbell <ian.campbell@citrix.com>

arm: fix build with gcc 4.7.x

As was already pointed out months ago (see
http://lists.xen.org/archives/html/xen-devel/2012-02/msg00826.html),
gcc 4.7.x (imo validly) refuses to take both -mcpu=cortex-a15 and
-march=armv7-a due to conflicting feature sets causing amibiguity in
instruction selection. Since the former implies the latter, just use
the former (and drop the -march=).

Signed-off-by: Jan Beulich <jbeulich@suse.com>
Acked-by: Ian Campbell <ian.campbell@citrix.com>

Fix check if guest enabled nested paging.

Fixes crashes with Windows guests when shadow-on-nested is used.

Signed-off-by: Christoph Egger <Christoph.Egger@amd.com>
Acked-by: Tim Deegan <tim@xen.org>

libxl: fix validation of scheduling parameters for sedf

2205914617cb does its job in correcting the "wrong domain being
considered" issue introduced by 9d1fd58ff602. Unfortunately, when
dealing (again!) with the sedf scheduler, it is required for the
vCPUs of a domain to have been allocated and setup already (in
the hypervisor), when the first call to libxl_domain_sched_params_get()
happens, and that is not true.

This fixes that by avoiding calling that function at all, as we
only need to know which scheduler the domain is running under,
and that is provided by libxl__domain_scheduler() which is safe
to be called there.

While at it, also improve a bit the comments about the whole
sedf parameter validation and mangling process.

Signed-off-by: Dario Faggioli <dario.faggioli@citrix.com>
Acked-by: Ian Campbell <ian.campbell@citrix.com>
Committed-by: Ian Campbell <ian.campbell@citrix.com>

x86/PCI: pass correct register value to XSM

When attempting to use AMD's extension to access the extended PCI config
space, only the low byte of the register number was being passed to XSM.
Include the correct value of the register if this feature is enabled;
otherwise, bits 24-30 of port cf8 are reserved, so disallow the invalid
access.

Signed-off-by: Daniel De Graaf <dgdegra@tycho.nsa.gov>
Don't fail the permission check except when the MSR can't be read.

Signed-off-by: Jan Beulich <jbeulich@suse.com>
Acked-by: Keir Fraser <keir@xen.org>
Committed-by: Jan Beulich <jbeulich@suse.com>

AMD IOMMU: add mechanism to protect their PCI devices' config spaces

Recent Dom0 kernels want to disable PCI MSI on all devices, yet doing
so on AMD IOMMUs (which get represented by a PCI device) disables part
of the functionality set up by the hypervisor.

Add a mechanism to mark certain PCI devices as having write protected
config spaces (both through port based [method 1] accesses and, for
x86-64, mmconfig), and use that for AMD's IOMMUs.

Note that due to ptwr_do_page_fault() being run first, there'll be a
MEM_LOG() issued for each such mmconfig based write attempt. If that's
undesirable, the order of the calls in fixup_page_fault() would need
to be swapped.

Signed-off-by: Jan Beulich <jbeulich@suse.com>
Tested-by: Wei Wang <wei.wang2@amd.com>
Acked-by: Keir Fraser <keir@xen.org>

libxl: validate scheduler parameters

This was previously done by xl itself however the domain was not
created at that point so there was no domid to check. This happened to
work on first boot because xl's global domid was initialised to zero
so we would (incorrectly) validate the new domain to be against
domain0. On reboot though we would try to use the old domain's id and
fail.

sched_params_valid is moved and gains a gc+domid parameters and
s/ctx/CTX/. The call is placed after
libxl__domain_build_info_setdefault in the create path, because
set_defaults doesn't have access to the domid and there are other
callers which don't even have a domid to give it.

Signed-off-by: Ian Campbell <ian.campbell@citrix.com>
Committed-by: Ian Campbell <ian.campbell@citrix.com>

x86-64: revert mmconfig part of c/s 24425:053a44894279

These additions did not fulfill their purpose - they checked hypervisor
config space accesses instead of guest (Dom0) ones.

Signed-off-by: Jan Beulich <jbeulich@suse.com>
Acked-by: Keir Fraser <keir@xen.org>

x86/PCI: fix guest_io_read() when pci_cfg_ok() denies access

For a multi-byte aligned read, this so far resulted in 0x00ff to be
put in the guest's register rather than 0xffff or 0xffffffff, which in
turn could confuse bus scanning functions (which, when reading vendor
and/or device IDs, expect to get back all zeroes or all ones).

As the value gets masked to the read width when merging back into the
full result, setting the initial value to all ones should not harm any
or the other cases.

Signed-off-by: Jan Beulich <jbeulich@suse.com>
Acked-by: Keir Fraser <keir@xen.org>

xl: fix sedf parameters checking

25468:9d1fd58ff602 was bogus in not letting a new domain being created
if its scheduling parameters --when running under the sedf scheduler--
were not fully specified, making creation fail like in this example
here below:

2012-06-16 07:37:47 Z executing ssh ... root@10.80.248.105 xl create /etc/xen/debian.guest.osstest.cfg
libxl: error: libxl.c:3619:sched_sedf_domain_set: setting domain sched sedf: Invalid argument
libxl: error: libxl_create.c:710:domcreate_bootloader_done: cannot (re-)build domain: -3
Parsing config from /etc/xen/debian.guest.osstest.cfg

This is due to the fact the values for period, slice, weight and
extratime should be consistent among each others, and if not all
are explicitly specified, someone has to make that happen. That
was right the purpose of the change in question, but it was failing
at achieving so.

This commit fixes things by forcing unspecified parameters to
sensible values, depending on the ones the user provided.

Signed-off-by: Dario Faggioli <dario.faggioli@citix.com>
Committed-by: Ian Jackson <ian.jackson@eu.citrix.com>

x86/mm: fix mod_l1_entry() return value when encountering r/o MMIO page

While putting together the workaround announced in
http://lists.xen.org/archives/html/xen-devel/2012-06/msg00709.html, I
found that mod_l1_entry(), upon encountering a set bit in
mmio_ro_ranges, would return 1 instead of 0 (the removal of the write
permission is supposed to be entirely transparent to the caller, even
more so to the calling guest).

Signed-off-by: Jan Beulich <jbeulich@suse.com>
Acked-by: Keir Fraser <keir@xen.org>

x86-64/EFI: document building and usage

Signed-off-by: Jan Beulich <jbeulich@suse.com>
Acked-by: Keir Fraser <keir@xen.org>

x86-64: don't allow non-canonical addresses to be set for any callback

Rather than deferring the detection of these to the point where they
get actually used (the fix for XSA-7, 25480:76eaf5966c05, causing a #GP
to be raised by IRET, which invokes the guest's [fragile] fail-safe
callback), don't even allow such to be set.

Signed-off-by: Jan Beulich <jbeulich@suse.com>
Acked-by: Keir Fraser <keir@xen.org>

x86, cpufreq: Change powernow's CPB status immediately

When command to modify turbo mode (CPB on AMD processors) comes
in the actual change happens later, when P-state transition is
requested. There is no time limit on when this transition will
occur and therefore change in CPB state may take long time from
the moment when command to toggle it is issued.

This patch makes CPB mode change happen immediately when request
is made.

Signed-off-by: Boris Ostrovsky <boris.ostrovsky@amd.com>
Committed-by: Keir Fraser <keir@xen.org>

libxl: propagate down the error from libxl_domain_sched_params_set

So that the caller (e.g., libxl__build_post() ) knows and can deal with it.

Signed-off-by: Dario Faggioli <dario.faggioli@citrix.com>
Acked-by: Ian Jackson <ian.jackson@eu.citrix.com>
Committed-by: Ian Jackson <ian.jackson@eu.citrix.com>

SVM: fix performance decrease with asid assignment

Do not clear asid cleanbit unconditionally. This shaves off 100 cycles
from the VMRUN instruction.

Signed-off-by: Christoph Egger <Christoph.Egger@amd.com>
Committed-by: Jan Beulich <jbeulich@suse.com>

x86-64: detect processors subject to AMD erratum #121 and refuse to boot

Processors with this erratum are subject to a DoS attack by unprivileged
guest users.

This is XSA-9 / CVE-2012-2934.

Signed-off-by: Jan Beulich <JBeulich@suse.com>
Signed-off-by: Ian Campbell <ian.campbell@citrix.com>
Committed-by: Ian Jackson <ian.jackson@eu.citrix.com>

x86_64: Do not execute sysret with a non-canonical return address

Check for non-canonical guest RIP before attempting to execute sysret.
If sysret is executed with a non-canonical value in RCX, Intel CPUs
take the fault in ring0, but we will necessarily already have switched
to the the user's stack pointer.

This is a security vulnerability, XSA-7 / CVE-2012-0217.

Signed-off-by: Jan Beulich <JBeulich@suse.com>
Signed-off-by: Ian Campbell <Ian.Campbell@citrix.com>
Signed-off-by: Ian Jackson <ian.jackson@eu.citrix.com>
Tested-by: Ian Campbell <Ian.Campbell@citrix.com>
Acked-by: Keir Fraser <keir.xen@gmail.com>
Committed-by: Ian Jackson <ian.jackson@eu.citrix.com>

x86: get rid of BOOT_TRAMPOLINE

We recently saw a machine that has the EBDA extending as low as
0x7c000, so that Xen fails to boot after relocating the trampoline.
To fix this, I removed BOOT_TRAMPOLINE and bootsym_phys completely.

Here are the parts:

1) the trampoline segment is set to 64k below the EBDA.  head.S grows
the ability to relocate the trampoline segment

2) reloc.c is made position-independent.  It allocates data below the
trampoline, whose address is passed in _eax.

3) cmdline.S is called before relocating, so all bootsym_phys there
become sym_phys.

Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
jb: - fall back to low memory size (instead of segment 0x7c00) if EBDA
      value is out of range
    - also add upper limit check on EBDA value
    - fix and simplify inline assembly operands in reloc_mbi_struct()
    - use lret instead of retf
    - renamed early_stack to wakeup_stack, defined and used now only
      in wakeup.S
    - aligned reloc.bin's end of .text to 16 bytes, so that checking
      __bss_start == end works reliably

Signed-off-by: Jan Beulich <jbeulich@suse.com>
Committed-by: Keir Fraser <keir@xen.org>

x86/nmi: Fix deadlock in unknown_nmi_error()

Additionally, correct the text description to reflect what is being
done, and make use of fatal_trap() in preference to kexec_crash() in
case an unknown NMI occurs before a kdump kernel has been loaded.

Signed-off-by: Andrew Cooper <andrew.cooper3@citrix.com>
Committed-by: Keir Fraser <keir@xen.org>

gdbsx: send virq to guest if gdbsx_vcpu_event is not active

gdbsx got broken along the way. During domain pause, don't send
VIRQ_DEBUGGER to guest if gdbsx is active on that guest.

Signed-off-by: Mukesh Rathor <mukesh.rathor@oracle.com>
Committed-by: Keir Fraser <keir@xen.org>

xl.cfg: document the cpuid= option

Signed-off-by: Olaf Hering <olaf@aepfle.de>
Committed-by: Ian Jackson <ian.jackson@eu.citrix.com>

tools/pygrub: Makefile cleanup

This patch removes the extra command `install pygrub` because this is already
done by the setup.py script.

Signed-off-by: Anthony PERARD <anthony.perard@citrix.com>
Acked-by: George Dunlap <george.dunlap@eu.citrix.com>
Committed-by: Ian Jackson <ian.jackson@eu.citrix.com>

tools/pygrub: Fix pygrub install

The script pygrub is currently installed in the wrong location. Instead of
/usr/lib/xen/bin, the script is installed in $destdir/usr/lib/xen/bin.
$(DESTDIR) is apply twice.

Signed-off-by: Anthony PERARD <anthony.perard@citrix.com>
Acked-by: George Dunlap <george.dunlap@eu.citrix.com>
Committed-by: Ian Jackson <ian.jackson@eu.citrix.com>
diff -r 474b5a6b1d91 -r 836e6618e86b tools/pygrub/Makefile
--- a/tools/pygrub/Makefile Fri Jun 08 16:39:24 2012 +0100
+++ b/tools/pygrub/Makefile Fri Jun 08 16:42:05 2012 +0100
@@ -12,7 +12,7 @@ build:
 install: all
  CC="$(CC)" CFLAGS="$(CFLAGS)" $(PYTHON) setup.py install \
  $(PYTHON_PREFIX_ARG) --root="$(DESTDIR)" \
- --install-scripts=$(DESTDIR)/$(PRIVATE_BINDIR) --force
+ --install-scripts=$(PRIVATE_BINDIR) --force
  $(INSTALL_PYTHON_PROG) src/pygrub $(DESTDIR)/$(PRIVATE_BINDIR)/pygrub
  $(INSTALL_DIR) $(DESTDIR)/var/run/xend/boot
  ln -sf $(PRIVATE_BINDIR)/pygrub $(DESTDIR)/$(BINDIR)

build: Correct typo introduced by c/s:25334

Signed-off-by: Juergen Gross <juergen.gross@ts.fujitsu.com>
Acked-by: Ian Campbell <ian.campbell@citrix.com>
Committed-by: Ian Jackson <ian.jackson@eu.citrix.com>

tools: disable libvchan build on NetBSD

NetBSD doesn't have a gntdev, so libvchan is unable to build due to
the lack of the header files gntalloc.h.

This is the error:

gcc  -O1 -fno-omit-frame-pointer -m64 -g -fno-strict-aliasing
 -std=gnu99 -Wall -Wstrict-prototypes -Wdeclaration-after-statement
 -D__XEN_TOOLS__ -MMD -MF .init.o.d -fno-optimize-sibling-calls
 -I../include -I.
 -I/root/xen/xen-netbsd/tools/libvchan/../../tools/xenstore
 -I/root/xen/xen-netbsd/tools/libvchan/../../tools/include
 -I/root/xen/xen-netbsd/tools/libvchan/../../tools/libxc
 -I/root/xen/xen-netbsd/tools/libvchan/../../tools/include  -c -o init.o init.c
 init.c:45:30: fatal error: xen/sys/gntalloc.h: No such file or directory
compilation terminated.
gmake[3]: *** [init.opic] Error 1
gmake[3]: *** Waiting for unfinished jobs....
init.c:45:30: fatal error: xen/sys/gntalloc.h: No such file or directory
compilation terminated.

Signed-off-by: Roger Pau Monne <roger.pau@citrix.com>
Acked-by: Christoph Egger <Christoph.Egger@amd.com>
Committed-by: Ian Jackson <ian.jackson@eu.citrix.com>

libxl: Store VNC passwd in xenstore with QEMU upstream.

Signed-off-by: Anthony PERARD <anthony.perard@citrix.com>
Committed-by: Ian Jackson <ian.jackson@eu.citrix.com>

libxc: do not "panic" if a kernel is not a bzImage.

Up until the point where we think this is a bzImage there is no point in
printing panicy messages -- some other loader will have a go (probably the
compressed ELF one)

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>

tools: readnote: Add bzImage kernel support

Add the check of bzImage kernel and make it work
with RHEL 6 big zImage kernel

Signed-off-by: Xuesen Guo <Xuesen.Guo@hitachiconsulting.com>
Acked-by: Ian Campbell <ian.campbell@citrix.com>
Committed-by: Ian Jackson <ian.jackson@eu.citrix.com>

xl: check for meaningful combination of sedf config file parameters

As it happens in the implementation of `xl sched-sedf -d ...', some
consistency checking is needed for the scheduling parameters when
they come from the config file.

Signed-off-by: Dario Faggioli <dario.faggioli@citrix.com>
Acked-by: Ian Jackson <ian.jackson@eu.citrix.com>
Acked-by: George Dunlap <george.dunlap@eu.citrix.com>
Committed-by: Ian Jackson <ian.jackson@eu.citrix.com>

QEMU_TAG update

tools: do not install qemu if just doing a build

Signed-off-by: David Vrabel <david.vrabel@citrix.com>
Acked-by: Ian Jackson <ian.jackson@eu.citrix.com>
Committed-by: Ian Jackson <ian.jackson@eu.citrix.com>

QEMU_TAG update

tools: pass EXTRA_CFLAGS via environment

Currently qemu-xen will be compiled with CFLAGS only if CFLAGS was
already in the environment during make invocation. If CFLAGS is in
environment then make will append all of the various flags specified in
xen Makefiles to this environment variable, which is then used in qemu
configure. Since qemu-xen is not ready for compiler flags like
"-std=gnu99" compilation will fail. If CFLAGS is not in environment,
then configure will use just its own "-O2 -g" because make does not
export its own CFLAGS variable.

>From a distro perspective, it is required to build libraries and
binaries with certain global cflags (arbitrary gcc options). Up to the
point when qemu-xen was imported it worked as expected by exporting
CFLAGS before 'make tools'.  Now qemu-upstream reuses these CFLAGS, but
it cant deal with the result.

This patch extends the tools Makefiles so that three new environment
variables are recognized:
  EXTRA_CFLAGS_XEN_TOOLS= specifies CFLAGS for the tools build.
  EXTRA_CFLAGS_QEMU_TRADITIONAL= specifies CFLAGS for old qemu.
  EXTRA_CFLAGS_QEMU_XEN= specifies CFLAGS for new qemu.

Special care needs to be taken in tools/firmware because the resulting
binaries are not linked with the hosts runtime libraries. These binaries
run in guest context. To avoid build errors from gcc options like
-fstack-protector, reuse existing practice to unset the new
EXTRA_CFLAGS_XEN_TOOLS for the firmware dirs.

The new feature can be used like this in a rpm xen.spec file:

   export EXTRA_CFLAGS_XEN_TOOLS="${RPM_OPT_FLAGS}"
   export EXTRA_CFLAGS_QEMU_TRADITIONAL="${RPM_OPT_FLAGS}"
   export EXTRA_CFLAGS_QEMU_XEN="${RPM_OPT_FLAGS}"
   ./configure \
    --libdir=%{_libdir} \
    --prefix=/usr
   make

Acked-by: Ian Jackson <ian.jackson@eu.citrix.com>
Committed-by: Ian Jackson <ian.jackson@eu.citrix.com>

xend: do not run a hotplug script from qemu on Linux

The current vif-hotplug-common.sh for renaming the tap device is failing
because it is racing with this script and therefore the device is unexpectedly
up when we come to rename it.

Fix this in the same way as libxl does, by disabling the script (only on
Linux).

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>

libxl: remove libxl__error_set prototype

The implementation went away in 25181:26f72d923cb9 "libxl: Crash (more
sensibly) on malloc failure".

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>

xenstore: set read_thread stacksize

xs_watch() creates a thread to wake watchers using default attributes. The
stacksize can be quite large (8 MB on Linux), applications that link against
xenstore end up having a larger memory footprint than necessary.

Signed-off-by: Simon Rowe <simon.rowe@eu.citrix.com>
Acked-by: Ian Jackson <ian.jackson@eu.citrix.com>
Committed-by: Ian Jackson <ian.jackson@eu.citrix.com>

xenstore: block signals in watch wakeup thread

The thread created to wakeup watchers is not intended to handle signals
(and a later patch will reduce it's stack size which makes it unsuitable
for doing so).

Signed-off-by: Simon Rowe <simon.rowe@eu.citrix.com>
Acked-by: Ian Jackson <ian.jackson@eu.citrix.com>
Committed-by: Ian Jackson <ian.jackson@eu.citrix.com>

xen: fix page_list_splice()

Other than in __list_splice(), the first element's prev pointer
doesn't need adjustment here - it already is PAGE_LIST_NULL. Rather
than fixing the assignment (to formally match __list_splice()), simply
assert that this assignment is really unnecessary.

Reported-by: Jisoo Yang <jisooy@gmail.com>
Signed-off-by: Jan Beulich <jbeulich@suse.com>
Also assert that the prev pointers are both PAGE_LIST_NULL.

Signed-off-by: Keir Fraser <keir@xen.org>
Committed-by: Keir Fraser <keir@xen.org>

xenpm, x86: Fix reporting of idle state average residency times

If CPU stays in the same idle state for the full duration of
xenpm sample then average residency may not be reported correctly
since usage counter will not be incremented.

In addition, in order to calculate averages correctly residence
time and usage counter should be read and written atomically.

Signed-off-by: Boris Ostrovsky <boris.ostrovsky@amd.com>
Committed-by: Keir Fraser <keir@xen.org>

xen: enhance dump_numa output

By showing the number of free pages on each node.

Signed-off-by: Dario Faggioli <dario.faggioli@citrix.com>
Acked-by: Ian Jackson <ian.jackson@eu.citrix.com>
Committed-by: Keir Fraser <keir@xen.org>

libxl: introduce LIBXL_DOMAIN_TYPE_INVALID

To avoid recent gcc complaining about:
libxl.c: In function 'libxl_primary_console_exec':
libxl.c:1233:9: error: case value '4294967295' not in enumerated type 'libxl_domain_type' [-Werror=switch]

Also:
 - have all the call sites of libxl__domain_type() return with error in
   case the function returns LIBXL_DOMAIN_TYPE_INVALID;
 - adjust all other code segments where -Wswitch makes would claim that
   LIBXL_DOMAIN_TYPE_INVALID is not handled by adding a "default: abort();"
   clause.

Signed-off-by: Dario Faggioli <dario.faggioli@citrix.com>
Signed-off-by: Christoph Egger <Christoph.Egger@amd.com>
Acked-by: Ian Campbell <ian.campbell@citrix.com>
[ ijc -- use LIBXL_DOMAIN_TYPE_INVALID instead of -1 for
         libxl_domain_build_info.type's keyvar_init_val.
      -- what used to be libxl_primary_console_exec is now in
         libxl__primary_console_find so resolve the rejected hunk
         accordingly ]
Committed-by: Ian Campbell <ian.campbell@citrix.com>

xl.cfg: document the maxmem= option

Signed-off-by: Olaf Hering <olaf@aepfle.de>
Acked-by: Ian Campbell <ian.campbell@citrix.com>
Committed-by: Ian Campbell <ian.campbell@citrix.com>

libxl: Add API to retrieve domain console tty

This api retrieve domain console from xenstore. With this new api, it is easy
to implement "virsh console" command in libvirt libxl driver.

Signed-off-by: Bamvor Jian Zhang <bjzhang@suse.com>
Acked-by: Ian Campbell <ian.campbell@citrix.com>
[ ijc -- use NOGC instead of 0 to allow improvements to this infrastructure
         in the future ]
Committed-by: Ian Campbell <ian.campbell@citrix.com>

libxl: convert malloc() to libxl__zalloc(NULL, ...)

And ditch the error handling in libxl_get_cpu_topology()

Signed-off-by: Dario Faggioli <dario.faggioli@citrix.com>
Acked-by: Ian Campbell <ian.campbell@citrix.com>
[ ijc -- define and use NOGC instead of NULL to allow improvements to this
          infrastructure in the future ]
Committed-by: Ian Campbell <ian.campbell@citrix.com>

ocaml: fix build after 25446:648508ee27a2, 25449:68d46c5ea0a3 et al.

These renamed a type and the associated functions and the ocaml bindings were
not updated to suit.

This also highlighted that libxl_domain_sched_params should not be DIR_IN since
it is also use as an output struct.

Signed-off-by: Ian Campbell <ian.campbell@citrix.com>
Committed-by: Ian Campbell <ian.campbell@citrix.com>

libxl: fix Makefile race bug relating to _paths.h

_paths.h needs to be in AUTOINCS.  That arranges for it to be an
explicit dependency of all object files.  This is necessary so that it
is made before any compilation is attempted.

Making it a dependency of xl.h (as in 25426:e53a1d3c212c) is harmless,
but not sufficient because that only takes effect if there is already
an autogenerated .d file naming xl.h as a dependency of relevant
object files.

Signed-off-by: Ian Jackson <ian.jackson@eu.citrix.com>
Acked-by: Ian Campbell <ian.campbell@citrix.com>
Committed-by: Ian Jackson <ian.jackson@eu.citrix.com>

libxl: fix typos in libxl_cpuid_parse_config

Fix typo in comment.
Fix cpuid_flags array init, use correct number of arguments for empty
array entry.

Signed-off-by: Olaf Hering <olaf@aepfle.de>
Acked-by: Ian Campbell <ian.campbell@citrix.com>
Committed-by: Ian Campbell <ian.campbell@citrix.com>

libxl: expose a single get/setter for domain scheduler parameters

This is consistent with having a single struct for all parameters.

Effectively renames and exports libxl__sched_set_params as
libxl_domain_sched_params_set. libxl_domain_sched_params_get is new.

Improve const correctness of the setters while I'm here.

Use shorter LOG macros when touching a line anyway.

Signed-off-by: Ian Campbell <ian.campbell@citrix.com>
Acked-by: George Dunlap <george.dunlap@eu.citrix.com>
Acked-by: Ian Jackson <ian.jackson@eu.citrix.com>
Committed-by: Ian Campbell <ian.campbell@citrix.com>

libxl: move libxl__sched_set_params into libxl.c

All the other sched functions are here and I'm just about to make those static
functions as I make libxl__sched_set_params the public function.

Signed-off-by: Ian Campbell <ian.campbell@citrix.com>
Acked-by: George Dunlap <george.dunlap@eu.citrix.com>
Acked-by: Ian Jackson <ian.jackson@eu.citrix.com>
Committed-by: Ian Campbell <ian.campbell@citrix.com>

libxl: make it possible to explicitly specify default sched params

To do so we define a discriminating value which is never a valid real value for
each parameter.

While there:

 - removed libxl_sched_*_domain in favour of libxl_domain_sched_params.
 - use this new functionality for the various xl commands which set sched
   parameters, which saves an explicit read-modify-write in xl.
 - removed call of xc_domain_getinfolist from a few functions which weren't
   actually using the result (looks like a cut and paste error)
 - fix xl which was setting period for a variety of different config keys.

Signed-off-by: Ian Campbell <ian.campbell@citrix.com>
Acked-by: George Dunlap <george.dunlap@eu.citrix.com>
Acked-by: Ian Jackson <ian.jackson@eu.citrix.com>
Committed-by: Ian Campbell <ian.campbell@citrix.com>

libxl: rename libxl_sched_params to libxl_domain_sched_params

Remove credit scheduler global options from the struct, they were never used
anyway.

Signed-off-by: Ian Campbell <ian.campbell@citrix.com>
Acked-by: George Dunlap <george.dunlap@eu.citrix.com>
Acked-by: Ian Jackson <ian.jackson@eu.citrix.com>
Committed-by: Ian Campbell <ian.campbell@citrix.com>

libxl: add internal function to get a domain's scheduler

This takes into account cpupools.

Add a helper to get the info for a single cpu pool, refactor libxl_list_cpupool
t use this. While there fix the leaks due to not disposing the partial list on
realloc failure in that function.

Fix the failure of sched_domain_output to free the poolinfo list.

Signed-off-by: Ian Campbell <ian.campbell@citrix.com>
Acked-by: George Dunlap <george.dunlap@eu.citrix.com>
Acked-by: Ian Jackson <ian.jackson@eu.citrix.com>
Committed-by: Ian Campbell <ian.campbell@citrix.com>

arm: Enable VFP at boot

Signed-off-by: Tim Deegan <tim@xen.org>
Acked-by: Ian Campbell <ian.campbell@citrix.com>
Committed-by: Ian Campbell <ian.campbell@citrix.com>

arm: allow ourselves access to all coprocessors in non-secure mode

We'll need it to be able to use the VFP extensions, for example.

Signed-off-by: Tim Deegan <tim@xen.org>
Acked-by: Ian Campbell <ian.campbell@citrix.com>
Committed-by: Ian Campbell <ian.campbell@citrix.com>

arm: missing __init annotation

Signed-off-by: Tim Deegan <tim@xen.org>
Acked-by: Ian Campbell <ian.campbell@citrix.com>
Committed-by: Ian Campbell <ian.campbell@citrix.com>

arm: avoid memory write in switch to Hyp mode

Assemble the new CPSR in registers instead.  It's slightly cleaner,
And makes it possible to have a read-only text section.

Signed-off-by: Tim Deegan <tim@xen.org>
Acked-by: Ian Campbell <ian.campbell@citrix.com>
Committed-by: Ian Campbell <ian.campbell@citrix.com>

arm: Move hyp-mode entry code out of line.

This code is grottier than the rest of the start-of-day code and
very specific to the software model we're developing on.
Segregate it accordingly, by putting it in its own file.

Signed-off-by: Tim Deegan <tim@xen.org>
Acked-by: Ian Campbell <ian.campbell@citrix.com>
Committed-by: Ian Campbell <ian.campbell@citrix.com>

arm: More interrupt setup at start-of-day for secondary CPUs

Signed-off-by: Tim Deegan <tim@xen.org>
Acked-by: Ian Campbell <ian.campbell@citrix.com>
Committed-by: Ian Campbell <ian.campbell@citrix.com>

arm: implement event injection

Implement vcpu_mark_events_pending using the vgic to inject PPI 31, that
we reserve for Xen usage.
In the future the interrupt used for event injection might be dynamic
and could be written into the device tree.
Otherwise it could be an SGI choosen by the guest and passed to Xen
through an hypercall.

Considering that:

- it is easy to determine if an event notification
interrupt has already been EOI'd by the guest just looking at the
evtchn_upcall_pending bit in the shared_info page;

- we can safely assume that there is at most one event notification
interrupt pending at any time in any set of LR registers because we
never inject more than a single event notification interrupt in one vcpu
(see vcpu_mark_events_pending);

we can avoid requesting maintenance interrupts for
VGIC_IRQ_EVTCHN_CALLBACK, provided that we check for event notification
interrupts that need to be cleared in the following places:

- maintenance interrupt entry;

- gic_set_guest_irq;

that is every time we are about to write to an LR.

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>

arm: remove VGIC_SOFTIRQ

Instead of using a softirq to check whether we need to set the VI bit in
the HCR (IRQ injection in the guest), always check the lr_mask on
leave_hypervisor_tail.

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>

arm: implement flush_tlb_all_local and flush_tlb_local

Call flush_tlb_all_local from create_p2m_entries after removing a page
from the p2m.

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>

arm: shared_info page allocation and mapping

Allocate the shared_info page at domain creation.

Implement arch_memory_op, only for XENMEM_add_to_physmap with space ==
XENMAPSPACE_shared_info, so that the guest can map the shared_info page.

Signed-off-by: Stefano Stabellini <stefano.stabellini@eu.citrix.com>
Signed-off-by: Ian Campbell <ian.campbell@citrix.com>
Committed-by: Ian Campbell <ian.campbell@citrix.com>

arm: replace list_del and INIT_LIST_HEAD with list_del_init

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>

arm: support fewer LR registers than virtual irqs

If the vgic needs to inject a virtual irq into the guest, but no free
LR registers are available, add the irq to a list and return.
Whenever an LR register becomes available we add the queued irq to it
and remove it from the list.
We use the gic lock to protect the list and the bitmask.

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>

passthrough: fix xsm-related oversight

Presumably a copy-and-paste mistake, which I also didn't notice while
unifying x86's and ia64's respective domctl implementations.

Signed-off-by: Jan Beulich <jbeulich@suse.com>
Acked-by: Keir Fraser <keir@xen.org>

x86/EDD: check MBR for BIOS magic before considering signature valid

Signed-off-by: Jan Beulich <JBeulich@suse.com>
Acked-by: Keir Fraser <keir@xen.org>

x86/amd: re-enable CPU topology extensions in case BIOS has disabled it

BIOS will switch off the corresponding feature flag on family
15h models 10h-1fh non-desktop CPUs.

The topology extension CPUID leafs are required to detect which
cores belong to the same compute unit. (thread siblings mask is
set accordingly and also correct information about L1i and L2
cache sharing depends on this).

W/o this patch we wouldn't see which cores belong to the same
compute unit and also cache sharing information for L1i and L2
would be incorrect on such systems.

Signed-off-by: Andreas Herrmann <andreas.herrmann3@amd.com>
Signed-off-by: Jan Beulich <jbeulich@suse.com>
Acked-by: Keir Fraser <keir@xen.org>
Committed-by: Jan Beulich <jbeulich@suse.com>

Rename: inslen -> insn_len

Signed-off-by: Keir Fraser <keir@xen.org>

tools/tests/xen-access: INT3 trap type is s/w exception.

Signed-off-by: Keir Fraser <keir@xen.org>

Add instruction length and trap type in HVMOP_inject_trap.

Signed-off-by: Xudong Hao <xudong.hao@intel.com>
Signed-off-by: Xiantao Zhang <xiantao.zhang@intel.com>
Signed-off-by: Keir Fraser <keir@xen.org>

xl: xl.h depends on geenrated file _paths.h

Fixes build error.

Signed-off-by: Ian Campbell <ian.campbell@citrix.com>
Committed-by: Ian Jackson <ian.jackson@eu.citrix.com>

hvm: Add instruction length to hvm_trap info.

Add documentation notes to vmx_inject_trap() and respect instruiction
length parameter for software-generated exceptions/interrupts.

Signed-off-by: Xudong Hao <xudong.hao@intel.com>
Signed-off-by: Eddie Dong <eddie.dong@intel.com>
Signed-off-by: Xiantao Zhang <xiantao.zhang@intel.com>
Signed-off-by: Keir Fraser <keir@xen.org>

xen: Define new struct hvm_trap and cleanup vmx exception

Define new struct hvm_trap to represent information of trap, and
renames hvm_inject_exception to hvm_inject_trap, then define a couple
of wrappers around that function for existing callers.

Signed-off-by: Keir Fraser <keir@xen.org>
Signed-off-by: Xudong Hao <xudong.hao@intel.com>
Committed-by: Keir Fraser <keir@xen.org>

xenpm: Fix reporting of C0 residence times

Idle state residence times as provided by pmstat_get_cx_stat() are not
reported precisely since remote core may be in idle state and
therefore has not updated its statistics at the time local core
collected them.  This causes C0 residencies as calculated by xenpm to
sometimes become negative.

Signed-off-by: Boris Ostrovsky <boris.ostrovsky@amd.com>
Committed-by: Keir Fraser <keir@xen.org>

x86: document register for 6th hypercall argument

From: David Vrabel <david.vrabel@citrix.com>

Signed-off-by: David Vrabel <david.vrabel@citrix.com>
Committed-by: Keir Fraser <keir@xen.org>

gnttab: cleanup

- introduce local variables (shortcuts for frequently used
  <dom>->grant_table)
- adjust first parameter of mapcount()
- drop lock acquisition from gnttab_get_version()
- remove hard tabs and adjust formatting

Signed-off-by: Jan Beulich <jbeulich@suse.com>
Tested-by: Andrew Thomas <andrew.thomas@oracle.com>
Committed-by: Keir Fraser <keir@xen.org>

gnttab: mark maptrack free list tail with an explicit terminator

... instead of using the mutable current limit.

This also addresses an apparent off-by-one mistake when checking for
exhaustion of the maptrack table.

Signed-off-by: Jan Beulich <jbeulich@suse.com>
Tested-by: Andrew Thomas <andrew.thomas@oracle.com>
Committed-by: Keir Fraser <keir@xen.org>

gnttab: don't use domain lock for serialization

Instead use the affected domain's grant table lock, at once reducing
the scopes during which locks are being held and hence allowing
significantly better parallelism.

Signed-off-by: Jan Beulich <jbeulich@suse.com>
Tested-by: Andrew Thomas <andrew.thomas@oracle.com>
Committed-by: Keir Fraser <keir@xen.org>

x86: fix nested HVM initialization

- no need for calling nestedhvm_setup() explicitly (can be a normal
  init-call, and can be __init)
- calling _xmalloc() for multi-page, page-aligned memory regions is
  inefficient - use alloc_xenheap_pages() instead
- albeit an allocation error is unlikely here, add error handling
  nevertheless (and have nestedhvm_vcpu_initialise() bail if an error
  occurred during setup)
- nestedhvm_enabled() must no access d->arch.hvm_domain without first
  checking that 'd' actually represents a HVM domain

Signed-off-by: Jan Beulich <JBeulich@suse.com>
Committed-by: Keir Fraser <keir@xen.org>

libxl: make libxl__e820_alloc a static function

Signed-off-by: Stefano Stabellini <stefano.stabellini@eu.citrix.com>
Acked-by: Ian Jackson <Ian.Jackson@eu.citrix.com>
Acked-by: Ian Campbell <Ian.Campbell@citrix.com>
Committed-by: Ian Campbell <ian.campbell@citrix.com>

libxl: move e820_names, e820_sanitize, libxl__e820_alloc to libxl_x86.c

Signed-off-by: Stefano Stabellini <stefano.stabellini@eu.citrix.com>
Acked-by: Ian Jackson <Ian.Jackson@eu.citrix.com>
Acked-by: Ian Campbell <Ian.Campbell@citrix.com>
Committed-by: Ian Campbell <ian.campbell@citrix.com>

libxl: Introduce libxl__arch_domain_create (x86 version)

Signed-off-by: Stefano Stabellini <stefano.stabellini@eu.citrix.com>
Acked-by: Ian Jackson <Ian.Jackson@eu.citrix.com>
Acked-by: Ian Campbell <Ian.Campbell@citrix.com>
Committed-by: Ian Campbell <ian.campbell@citrix.com>

libxl: Introduce libxl__arch_domain_create

Introduce an arch specific internal domain creation function.
The X86 version of libxl__arch_domain_create is going to be introduced
in the next few patches, to make it easier to spot all the code motions.

Signed-off-by: Stefano Stabellini <stefano.stabellini@eu.citrix.com>
Acked-by: Ian Jackson <Ian.Jackson@eu.citrix.com>
Acked-by: Ian Campbell <Ian.Campbell@citrix.com>
Committed-by: Ian Campbell <ian.campbell@citrix.com>

arm: compile libxl

libxl_cpuid_destroy has been renamed to libxl_cpuid_dispose; also cpuid
functions are only available on x86, so move them to libxl_cpuid.c.

Signed-off-by: Stefano Stabellini <stefano.stabellini@eu.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>

arm: compile xentrace

Signed-off-by: Stefano Stabellini <stefano.stabellini@eu.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>

arm: compile memshr

Signed-off-by: Stefano Stabellini <stefano.stabellini@eu.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>