x86: move both exception tables into .rodata

While they are being written during early boot (when sorting them),
that writing takes place before we actually start fiddling with page
table permissions, so these tables can benefit from getting write
protected just like ordinary r/o data does (for now only when using
2M mappings).

Signed-off-by: Jan Beulich <jbeulich@suse.com>
Reviewed-by: Andrew Cooper <andrew.cooper3@citrix.com>

x86: partially revert use of 2M mappings for hypervisor image

As explained by Andrew in
http://lists.xenproject.org/archives/html/xen-devel/2016-03/msg01380.html
that change makes the uncompressed xen.gz image too large for certain
boot environments. As a result this change makes some of the effects of
commits cf393624ee ("x86: use 2M superpages for text/data/bss
mappings") and 53aa3dde17 ("x86: unilaterally remove .init mappings")
conditional, restoring alternative previous code where necessary. This
is so that xen.efi can still benefit from the new mechanisms, as it is
unaffected by said limitations.

Signed-off-by: Jan Beulich <jbeulich@suse.com>
Reviewed-by: Andrew Cooper <andrew.cooper3@citrix.com>

sched RTDS: use uint64_t for tracing time values

such as deadline and budget. Packing is necessary to make
it possible for xentrace_format to properly interpreet the
records.

Signed-off-by: Dario Faggioli <dario.faggioli@citrix.com>
Reviewed-by: Meng Xu <mengxu@cis.upenn.edu>

vmx: restore debug registers when injecting #DB traps

Commit a929bee0e652 ("x86/vmx: Fix injection of #DB traps following
XSA-156") prevents an infinite loop in certain #DB traps. However, it
changed the behavior to not call hvm_hw_inject_trap() for #DB and #AC
traps which which means that the debug registers are not restored
correctly and nullified commit b56ae5b48c38 ("VMX: fix/adjust trap
injection").

To fix this, restore the original code path through hvm_inject_trap(),
but ensure that the struct hvm_trap is populated with all the required
data.

Signed-off-by: Ross Lagerwall <ross.lagerwall@citrix.com>
Reviewed-by: Jan Beulich <jbeulich@suse.com>
Acked-by: Kevin Tian <kevin.tian@intel.com>

restore p2m_access_t enum order to allow bitmask semantics

Nested hap code assumed implict bitmask semantics of the p2m_access_t
enum prior to C/S 4c63692d7c38c5ac414fe97f8ef37b66e05abe5c

The change to the enum ordering broke this assumption and caused functional
problems for the nested hap code. As it may be error prone to audit and find
all other p2m_access users assuming bitmask semantics, instead restore the
previous enum order and make it explict that bitmask semantics are to be
preserved for the read, write and execute access types.

Signed-off-by: Malcolm Crossley <malcolm.crossley@citrix.com>
Reviewed-by: Andrew Cooper <andrew.cooper3@citrix.com>
Reviewed-by: George Dunlap <george.dunlap@citrix.com>

libxl: use LIBXL__LOG in libxl_ctx_alloc

Coverity points out that using LOG macro dereferences gc which is NULL
at that point. Use LIBXL__LOG instead.

CID: 1343291

Signed-off-by: Wei Liu <wei.liu2@citrix.com>
Acked-by: Ian Jackson <ian.jackson@eu.citrix.com>

x86emul: fix 32-bit test build

Commit 5644ce0142 ("x86emul: relax asm() constraints") introduced a
64-bit only instruction suffix, which breaks running the emulator test
on a 32-bit system. Mirror __OS (and _OP for completeness) to the test
wrapper source file.

Signed-off-by: Jan Beulich <jbeulich@suse.com>
Reviewed-by: Andrew Cooper <andrew.cooper3@citrix.com>

x86emul: check host features alongside guest ones where needed

Signed-off-by: Jan Beulich <jbeulich@suse.com>
Reviewed-by: Andrew Cooper <andrew.cooper3@citrix.com>

x86: rename XMM* features to SSE*

The latter are their canonical names, used already in the instruction
emulator.

Signed-off-by: Jan Beulich <jbeulich@suse.com>
Reviewed-by: Andrew Cooper <andrew.cooper3@citrix.com>

IOMMU: make the pcidevs_lock a recursive one

The pcidevs_lock is going to be recursively taken for hiding ATS
device, when VT-d Device-TLB flush timed out.

Signed-off-by: Quan Xu <quan.xu@intel.com>
Acked-by: Kevin Tian <kevin.tian@intel.com>
Reviewed-by: Dario Faggioli <dario.faggioli@citrix.com>
pcidevs_is_locked() -> pcidevs_locked(). Add __must_check to
pcidevs_locked() and pcidevs_trylock() declarations.

Signed-off-by: Jan Beulich <jbeulich@suse.com>

AMD IOMMU: fix an init time spinlock flaw

pcidevs_lock doesn't require interrupts to be disabled while being acquired.
However there remains an exception in AMD IOMMU code, where the lock is
acquired with interrupt disabled. This inconsistency might lead to deadlock.

The fix is straightforward to use spin_lock instead. Also interrupt has been
enabled when this function is invoked, so we're sure consistency around
pcidevs_lock can be guaranteed after this fix.

Signed-off-by: Quan Xu <quan.xu@intel.com>
Reviewed-by: Dario Faggioli <dario.faggioli@citrix.com>
Reviewed-by: Kevin Tian <kevin.tian@intel.com>
Acked-by: Jan Beulich <jbeulich@suse.com>

libxl: ensure var is inited in libxl__domain_firmware

Some versions of GCC complain that the 'firmware' variable can be used
uninitialized. It looks like the switch inside of the else case is just
confusing GCC.

Signed-off-by: Doug Goldstein <cardoe@cardoe.com>
Acked-by: Wei Liu <wei.liu2@citrix.com>

libxc: simplify error handling in meminit_hvm

The hunk that prints out error message can only be reached from the loop
that allocates memory. Move that hunk into the loop to avoid confusing
gcc.

Reported-by: Doug Goldstein <cardoe@cardoe.com>
Signed-off-by: Wei Liu <wei.liu2@citrix.com>
Reviewed-by: Doug Goldstein <cardoe@cardoe.com>
Acked-by: Roger Pau Monne <roger.pau@citrix.com>

sched: add Meng as RTDS maintainer

Meng Xu is one of the maintainers of the RT-Xen project,
which is from where the RTDS scheduler comes. He also
is the main author of the version of RTDS that we currently
have here upstream.

Since the upstreaming effort, he's continued looking after
the code, engaging with the community, coming and presenting
at XenSummits and, last but not least, doing development
himself as well as directing the work of others, with the
aim of improving the scheduler.

In summary, he has reached the point, both from thechnical
and community engagement point of views, where he can
effectively serve as a maintainer of RTDS code.

Signed-off-by: Dario Faggioli <dario.faggioli@citrix.com>
Acked-by: Meng Xu <mengxu@cis.upenn.edu>
Acked-by: George Dunlap <george.dunlap@citrix.com>

x86: don't flush the whole cache when changing cachability

Introduce the FLUSH_VA_VALID flag to flush_area_mask() and friends to
say that it is safe to use CLFLUSH (i.e., the virtual address is still
valid).

Use this when changing the cachability of the Xen direct mappings (in
response to the guest changing the cachability of its mappings). This
significantly improves performance by avoiding an expensive WBINVD.

This fixes a performance regression introduced by
c61a6f74f80eb36ed83a82f713db3143159b9009 (x86: enforce consistent
cachability of MMIO mappings), the fix for XSA-154.

e.g., A set_memory_wc() call in Linux:

before: 4097 us
after:    47 us

Signed-off-by: David Vrabel <david.vrabel@citrix.com>
Reviewed-by: Jan Beulich <jbeulich@suse.com>

x86/shadow: avoid extra local array variable

mfns[2] was there just because struct sh_emulate_ctxt's two MFN values
can't be used to hand to vmap(). Making the structure fields an array
avoids the extra copying.

Signed-off-by: Jan Beulich <jbeulich@suse.com>
Reviewed-by: Andrew Cooper <andrew.cooper3@citrix.com>
Reviewed-by: Tim Deegan <tim@xen.org>

x86/shadow: compile most write emulation code just once

No need to compile all of this code three times, as most of it really
is guest mode independent. The savings are between 3k and 4k of binary
code in my builds. For this to fully work out, the sh_gva_to_gfn()
calls are being replaced by indirect ones through the paging mode
table.

No functional change (i.e. only formatting and naming changes) except
for
- sh_emulate_map_dest()'s user mode check corrected for the PV case
  (affecting debugging mode only, this isn't being split out)
- simplifying the vaddr argument to emulate_gva_to_mfn() for the second
  part in the cross page write case

Signed-off-by: Jan Beulich <jbeulich@suse.com>
Reviewed-by: Tim Deegan <tim@xen.org>
Reviewed-by: Andrew Cooper <andrew.cooper3@citrix.com>

x86/HVM: don't disable the REP emulation optimizations for regular IO

Currently REP emulations optimizations remain disabled even if
the emulation does not happen as a result of a vm_event reply
requestion emulation (i.e. even for regular IO). This patch takes
emulate_each_rep into account only if emulation has been requested
by a vm_event-capable application, and is a noticeable speed
optimization for monitored guests.

Signed-off-by: Razvan Cojocaru <rcojocaru@bitdefender.com>
Reviewed-by: Andrew Cooper <andrew.cooper3@citrix.com>
Reviewed-by: Tamas K Lengyel <tamas@tklengyel.com>
Acked-by: Jan Beulich <jbeulich@suse.com>

x86/xstate: undo bogus adjustment to xsave()

This reverts an unintended change in commit 879b44b041 ("x86/fpu: add
a per-domain field to set the width of FIP/FDP"), which I had done
intermediately while fixing the build issue: After having reverted that
adjustment I must have forgotten to "git add" the adjustment.

Signed-off-by: Jan Beulich <jbeulich@suse.com>
Reviewed-by: Andrew Cooper <andrew.cooper3@citrix.com>

tools/misc: drop unused function from gtracestat

Fixes a build failure with Clang.

Signed-off-by: Andrew Cooper <andrew.cooper3@citrix.com>
Acked-by: Wei Liu <wei.liu2@citrix.com>

tools/foreign: avoid using alignment directives when not appropriate

The foreign header generation blindly replaces 'uint64_t' with '__align8__
uint64_t', to get correct alignment when built as 32bit.  This is correct in
most circumstances, but Clang objects to two specific uses.

 * Inside a sizeof() expression
 * As part of a typecast

An example error looks like:

/local/xen.git/tools/libxc/../../tools/include/xen/foreign/x86_64.h:204:44:
error: 'aligned' attribute ignored when parsing type [-Werror,-Wignored-attributes]
    __align8__ uint64_t evtchn_mask[sizeof(__align8__ uint64_t) * 8];
                                           ^~~~~~~~~~
/local/xen.git/tools/libxc/../../tools/include/xen/foreign/x86_64.h:13:36:
note: expanded from macro '__align8__'
                                   ^~~~~~~~~~~

This sedary is sufficient to fix all the bad examples without touching any of
the legitimate uses, and is more simple than teaching mkheader.py how to parse
C.

Signed-off-by: Andrew Cooper <andrew.cooper3@citrix.com>
Reviewed-by: Jan Beulich <jbeulich@suse.com>
Acked-by: Wei Liu <wei.liu2@citrix.com>

add Julien as ARM maintainer

I would like to propose Julien Grall (julien.grall@arm.com) as
co-maintainer for Xen on ARM. His track record of contributions to the
project is outstanding and I think speaks for itself.

Julien made multiple public presentations about Xen on ARM at Xen
Developer Summit and Linaro Connect conferences. He led the development
of several key features, such as SMMU support and non-PCI passthrough
and participated in the design and the review of large code
contributions from others.

Signed-off-by: Stefano Stabellini <stefano.stabellini@eu.citrix.com>
Acked-by: Julien Grall <julien.grall@arm.com>
Acked-by: Ian Jackson <ian.jackson@eu.citrix.com>
Acked-by: Jan Beulich <jbeulich@suse.com>

mm: fix page_list_* helpers to evaluate all their arguments

If an architecture does not provide a custom page_list_entry, default
page_list_* helpers are provided, wrapping list_head as an underlying type for
page_list_head.

The two declarations of the page_list_* helpers differ between defines and
static inline functions, where the defines discard some of their parameters.

This causes a compilation failure if CONFIG_BIGMEM and debug=n in p2m-pod.c:

  p2m-pod.c: In function \91p2m_pod_cache_add\92:
  p2m-pod.c:72:20: error: unused variable \91d\92 [-Werror=unused-variable]
       struct domain *d = p2m->domain;
                      ^
  cc1: all warnings being treated as errors

because the use of d outside of the !NDEBUG section doesn't get evaluated as a
parameter by page_list_del().

Fix this by turning all #defines into static inline functions, so all
parameters are evaluated even if they are not used.

While editing this area, correct the return type of page_list_empty from int
to bool_t.

No functional change.

Reported-by: Doug Goldstein <cardoe@cardoe.com>
Signed-off-by: Andrew Cooper <andrew.cooper3@citrix.com>
Reviewed-by: Doug Goldstein <cardoe@cardoe.com>
Acked-by: George Dunlap <george.dunlap@eu.citrix.com>
Reviewed-by: Jan Beulich <jbeulich@suse.com>

mm: introduce arch_free_heap_page()

common/page_alloc.c references d->arch.relmem_list, which only exists on x86.
This only compiles on ARM because page_list_del2() discards its second
argument.

Introduce a new common arch_free_heap_page() which only uses common lists in
struct domain, and allow an architecture to override this with a custom
alternative.  x86 then provides a custom arch_free_heap_page() which takes
care of managing d->arch.relmem_list.

No functional change.

Signed-off-by: Andrew Cooper <andrew.cooper3@citrix.com>
Acked-by: Jan Beulich <jbeulich@suse.com>

x86/alternatives: correct near branch check

Make sure the near JMP/CALL check doesn't consume uninitialized
data, not even in a benign way. And relax the length check at once.

Signed-off-by: Jan Beulich <jbeulich@suse.com>
Reviewed-by: Andrew Cooper <andrew.cooper3@citrix.com>

domctl: add force flag to xen_domctl_vcpuaffinity for undoing pin override

Add a XEN_VCPUAFFINITY_FORCE flag to xen_domctl_vcpuaffinity structure
which will allow to undo a SCHEDOP_pin_override in case of a driver
error of the hardware domain which didn't do the expected
SCHEDOP_pin_override with cpu < 0 which would have done the undo
operation.

Signed-off-by: Juergen Gross <jgross@suse.com>
Acked-by: Jan Beulich <jbeulich@suse.com>

sched: add hypercall option to override and restore vcpu affinity

Some hardware (e.g. Dell studio 1555 laptops) require SMIs to be
called on physical cpu 0 only. Linux drivers like dcdbas or i8k try
to achieve this by pinning the running thread to cpu 0, but in Dom0
this is not enough: the vcpu must be pinned to physical cpu 0 via
Xen, too.

Add a stable hypercall option SCHEDOP_pin_override to the sched_op
hypercall to achieve this. It is taking a physical cpu number as
parameter. If pinning is possible (the calling domain has the
privilege to make the call and the cpu is available in the domain's
cpupool) the calling vcpu is pinned to the specified cpu. The old
cpu affinity is saved. To undo the override pinning a negative cpu
value is specified. This will restore the original cpu affinity of
the vcpu.

Signed-off-by: Juergen Gross <jgross@suse.com>
Acked-by: Dario Faggioli <dario.faggioli@citrix.com>
Reviewed-by: Jan Beulich <jbeulich@suse.com>

cpupool: correct error handling when removing cpu from cpupool

When schedule_cpu_switch() called from cpupool_unassign_cpu_helper()
returns an error, the domlist_read_lock isn't released again.

As cpu_disable_scheduler() might have changed affinity of some
domains domain_update_node_affinity() must be called for all domains
in the cpupool even in error case.

Even if looking weird it is okay to let the to be removed cpu set in
cpupool_free_cpus in case of an error returned by
cpu_disable_scheduler(). Add a comment explaining the reason for this.

Signed-off-by: Juergen Gross <jgross@suse.com>
Acked-by: Dario Faggioli <dario.faggioli@citrix.com>

x86/vPMU: do not clobber IA32_MISC_ENABLE

The VMX RDMSR intercept for MSR_IA32_MISC_ENABLE falls through into
vpmu_do_rdmsr(), so that core2_vpmu_do_rdmsr() may play with the PTS and PEBS
UNAVAIL bits.

Some 64bit Windows include IA32_MISC_ENABLE in the set of items checked by
PatchGuard, and will suffer a BSOD 0x109 CRITICAL_STRUCTURE_CORRUPTION if the
contents change on migrate.

The vPMU infrastructure should not clobber IA32_MISC_ENABLE at all.

Signed-off-by: Andrew Cooper <andrew.cooper3@citrix.com>
Reviewed-by: Boris Ostrovsky <boris.ostrovsky@oracle.com>

hvmloader: use xen/errno.h rather than the host systems errno.h

hvmloader is unhosted, and shouldn't use the system errno.h.  It already has
to use Xen's errno.h for other hypercalls.  The use of public/io/xs_wire.h
requires the use of un-prefixed errno values.

This fixes the build on stricter toolchains where requesting -fno-builtin does
reduce the include path as much as it can.

Reported-by: Doug Goldstein <cardoe@cardoe.com>
Signed-off-by: Andrew Cooper <andrew.cooper3@citrix.com>
Reviewed-by: Doug Goldstein <cardoe@cardoe.com>
Reviewed-by: Wei Liu <wei.liu2@citrix.com>
Acked-by: Jan Beulich <jbeulich@suse.com>

public/errno: Reduce complexity of inclusion

The inclusion rules conditions for errno.h were unnecesserily complicated, and
required the includer to jump through hoops if they wished to avoid getting
multiple namespaces worth of constants.

Simply the logic, and document what is going on.

Signed-off-by: Andrew Cooper <andrew.cooper3@citrix.com>
Acked-by: Jan Beulich <jbeulich@suse.com>

Merge branch 'tracing/sched-events-improvements' of git://xenbits.xen.org/people/dariof/xen into staging

* 'tracing/sched-events-improvements' of git://xenbits.xen.org/people/dariof/xen:
  xenalyze: handle Credit2 scheduler events
  xenalyze: handle Credit1 scheduler events
  xenalyze: handle scheduling events
  xentrace: formats: add domain create and destroy events.
  xentrace: formats: add events from RTDS scheduler
  xentrace: formats: add events from Credit2 scheduler
  xentrace: formats: add events from Credit scheduler
  xentrace: formats: update format of scheduling events

On the mailing list all patches have:
Acked-by: George Dunlap <george.dunlap@citrix.com>
and have been
Reviewed-by: Konrad Rzeszutek Wilk <konrad.wilk@oracle.com>
so in they go.

Signed-off-by: Konrad Rzeszutek Wilk <konrad.wilk@oracle.com>

travis: actually disable debug for non-debug

Non-debug builds need to explicitly disable debug due to debug being
defaulted to y in Config.mk

(Xen keeps debug=y in the staging branch. When Xen is in rc-X stage
the debug is altered so that the builds don't have the debug
option enabled by default).

Signed-off-by: Doug Goldstein <cardoe@cardoe.com>
Reviewed-by: Konrad Rzeszutek Wilk <konrad.wilk@oracle.com>

travis: use matching C++ for GCC version

When we use GCC 5.x, we need to install the C++ compiler and the C
compiler together because QEMU tests for feature flags against the C
compiler and assumes the C++ compiler has them. We also have to
ensure that GCC C++ is used. Have to do the modification of the CXX variable
in two steps to ensure we support older versions of bash in use by the
test machines. While we're at it simply how we select our compiler.

Signed-off-by: Doug Goldstein <cardoe@cardoe.com>
Reviewed-by: Konrad Rzeszutek Wilk <konrad.wilk@oracle.com>

travis: skip building coverity, smoke, and master

Skip building of the coverity, smoke, stable, and master branches since
they just fast forward from staging.

Suggested-by: Andrew Cooper <andrew.cooper3@citrix.com>
Signed-off-by: Doug Goldstein <cardoe@cardoe.com>
Reviewed-by: Konrad Rzeszutek Wilk <konrad.wilk@oracle.com>

xenalyze: handle Credit2 scheduler events

so the trace will show properly decoded info,
rather than just a bunch of hex codes.

Signed-off-by: Dario Faggioli <dario.faggioli@citrix.com>
Reviewed-by: Konrad Rzeszutek Wilk <konrad.wilk@oracle.com>
---
Cc: George Dunlap <george.dunlap@eu.citrix.com>
Cc: Ian Jackson <ian.jackson@eu.citrix.com>
Cc: Ian Campbell <ian.campbell@citrix.com>
Cc: Wei Liu <wei.liu2@citrix.com>
Cc: Olaf Hering <olaf@aepfle.de>
Cc: Konrad Rzeszutek Wilk <konrad.wilk@oracle.com>
---
Changes from v1:
 * '} * r =' turned into '} *r =', as requested
   during review.

xenalyze: handle Credit1 scheduler events

so the trace will show properly decoded info,
rather than just a bunch of hex codes.

Signed-off-by: Dario Faggioli <dario.faggioli@citrix.com>
Reviewed-by: Konrad Rzeszutek Wilk <konrad.wilk@oracle.com>
---
Cc: George Dunlap <george.dunlap@eu.citrix.com>
Cc: Ian Jackson <ian.jackson@eu.citrix.com>
Cc: Ian Campbell <ian.campbell@citrix.com>
Cc: Wei Liu <wei.liu2@citrix.com>
Cc: Olaf Hering <olaf@aepfle.de>
---
Changes from v1:
 * '} * r =' turned into '} *r =', as requested
   during review.

xenalyze: handle scheduling events

so the trace will show properly decoded info,
rather than just a bunch of hex codes.

Signed-off-by: Dario Faggioli <dario.faggioli@citrix.com>
Reviewed-by: Konrad Rzeszutek Wilk <konrad.wilk@oracle.com>
---
Cc: George Dunlap <george.dunlap@eu.citrix.com>
Cc: Ian Jackson <ian.jackson@eu.citrix.com>
Cc: Ian Campbell <ian.campbell@citrix.com>
Cc: Wei Liu <wei.liu2@citrix.com>
Cc: Olaf Hering <olaf@aepfle.de>
---
Changes from v1:
 * SCHED_DOM_{ADD,REM} handling slightly changed, to avoid
   confusion with DOM0_DOM_{ADD,REM} (introduced later in
   the series);
 * '} * r =' turned into '} *r =', as requested
   during review.

xentrace: formats: add domain create and destroy events.

Signed-off-by: Dario Faggioli <dario.faggioli@citrix.com>
---
Cc: George Dunlap <george.dunlap@eu.citrix.com>
Cc: Ian Jackson <ian.jackson@eu.citrix.com>
Cc: Ian Campbell <ian.campbell@citrix.com>
Cc: Wei Liu <wei.liu2@citrix.com>
Cc: Olaf Hering <olaf@aepfle.de>
Cc: Konrad Rzeszutek Wilk <konrad.wilk@oracle.com>
---
Changes from v2:
 * new patch in the series.

xentrace: formats: add events from RTDS scheduler

Signed-off-by: Dario Faggioli <dario.faggioli@citrix.com>
---
Cc: George Dunlap <george.dunlap@eu.citrix.com>
Cc: Meng Xu <xumengpanda@gmail.com>
Cc: Tianyang Chen <tiche@seas.upenn.edu>
Cc: Ian Jackson <ian.jackson@eu.citrix.com>
Cc: Ian Campbell <ian.campbell@citrix.com>
Cc: Wei Liu <wei.liu2@citrix.com>
Cc: Olaf Hering <olaf@aepfle.de>

xentrace: formats: add events from Credit2 scheduler

Signed-off-by: Dario Faggioli <dario.faggioli@citrix.com>
---
Cc: George Dunlap <george.dunlap@eu.citrix.com>
Cc: Ian Jackson <ian.jackson@eu.citrix.com>
Cc: Ian Campbell <ian.campbell@citrix.com>
Cc: Wei Liu <wei.liu2@citrix.com>
Cc: Olaf Hering <olaf@aepfle.de>
---
Changes from v1:
 * fix typo in two events (rq_idx/rq_id)., as requested during
   review.

xentrace: formats: add events from Credit scheduler

Signed-off-by: Dario Faggioli <dario.faggioli@citrix.com>
Reviewed-by: Konrad Rzeszutek Wilk <konrad.wilk@oracle.com>
---
Cc: George Dunlap <george.dunlap@eu.citrix.com>
Cc: Ian Jackson <ian.jackson@eu.citrix.com>
Cc: Ian Campbell <ian.campbell@citrix.com>
Cc: Wei Liu <wei.liu2@citrix.com>
Cc: Olaf Hering <olaf@aepfle.de>

xentrace: formats: update format of scheduling events

to include the vcpu IDs, in a way that matches
how the "dom:vcpu" couple is displayed in other
events (runstate changes).

Also add the trace for TRC_SCHED_SHUTDOWN_CODE which
was missing and was done via SCHEDOP_shutdown_code hypercall.
(TRC_SCHED_SHUTDOWN trace was present).

Signed-off-by: Dario Faggioli <dario.faggioli@citrix.com>
Reviewed-by: Konrad Rzeszutek Wilk <konrad.wilk@oracle.com>
---
Cc: George Dunlap <george.dunlap@eu.citrix.com>
Cc: Ian Jackson <ian.jackson@eu.citrix.com>
Cc: Ian Campbell <ian.campbell@citrix.com>
Cc: Wei Liu <wei.liu2@citrix.com>
Cc: Olaf Hering <olaf@aepfle.de>
---
Changes from v1:
 * enhanced changelog, as suggested during review.

arm/timer: fix panic when booting with DT

While to support ACPI, patch "arm/acpi: Parse GTDT to initialize timer"
refactors the functions preinit_xen_time and init_xen_time. But it
wrongly moves the platform_get_irq from init_xen_time to
preinit_dt_xen_time and this will cause booting failure.

So move platform_get_irq back to init_xen_time to fix it.

Signed-off-by: Shannon Zhao <shannon.zhao@linaro.org>
Reviewed-by: Stefano Stabellini <stefano.stabellini@eu.citrix.com>

x86/HVM: re-format cache attribute pinning code

No intended functional change, albeit it includes ditching a redundant
is_hvm_domain().

Signed-off-by: Jan Beulich <jbeulich@suse.com>
Reviewed-by: Andrew Cooper <andrew.cooper3@citrix.com>

x86/HVM: adjust hvm_get_mem_pinned_cacheattr() GFN parameter

Make it gfn_t and rename it accordingly.

Suggested-by: Andrew Cooper <andrew.cooper3@citrix.com>
Signed-off-by: Jan Beulich <jbeulich@suse.com>
Reviewed-by: Andrew Cooper <andrew.cooper3@citrix.com>
Acked-by: Tim Deegan <tim@xen.org>

x86/HVM: limit flushing on cache attribute pinning adjustments

Avoid cache flush on EPT when removing a UC- range, since when used
this type gets converted to UC anyway (there's no UC- among the types
valid in MTRRs and hence EPT's emt field).

We might further wwant to consider only forcing write buffer flushes
when removing WC ranges.

Signed-off-by: Jan Beulich <jbeulich@suse.com>
Reviewed-by: Andrew Cooper <andrew.cooper3@citrix.com>

x86/HVM: adjust hvm_set_mem_pinned_cacheattr() error indications

Make it return an error on bad domain kind or obviously bad GFN range.

Signed-off-by: Jan Beulich <jbeulich@suse.com>
Reviewed-by: Andrew Cooper <andrew.cooper3@citrix.com>

x86/HVM: remove unnecessary indirection from hvm_get_mem_pinned_cacheattr()

Its return value can easily serve the purpose. We cannot, however,
return unspecific "success" anymore for a domain of the wrong type -
since no caller exists that would call this for PV domains, simply add
an ASSERT().

Signed-off-by: Jan Beulich <jbeulich@suse.com>
Reviewed-by: Andrew Cooper <andrew.cooper3@citrix.com>
Acked-by: Tim Deegan <tim@xen.org>

x86/HVM: honor cache attribute pinning for RAM only

Call hvm_get_mem_pinned_cacheattr() for RAM ranges only, and only when
the guest has a physical device assigned: XEN_DOMCTL_pin_mem_cacheattr
is documented to be intended for RAM only.

Signed-off-by: Jan Beulich <jbeulich@suse.com>
Reviewed-by: Andrew Cooper <andrew.cooper3@citrix.com>
Acked-by: Tim Deegan <tim@xen.org>

x86: use "unsigned int" for cache attribute values

... where suitable. But note that the type of "cacheattr" in
get_page_from_l1e() specifically needs to remain "unsigned long".

Signed-off-by: Jan Beulich <jbeulich@suse.com>
Reviewed-by: Andrew Cooper <andrew.cooper3@citrix.com>

public/io/netif.h: make control ring hash protocol more general

This patch modified the control ring protocol (of which there is
not yet an implementation) to make it more general. Most of the
concepts are not limited to toeplitz hashing so it's best not to
make them unnecessarily specific.

Apart from changing the names of various definitions and modifying
comments, this patch:

- Adds a new control message type to select a hash algorithm.
- Adds a reference implementation of the toeplitz hash.
- Changes the 'toeplitz' extra info fragment into a 'hash' extra
  info fragment and replaces the octet of padding with the index of
  the algorithm that was used to create the hash value.
- Relaxes the restriction that the mapping table has to be
  power-of-2 sized.

The patch also fixes a few spelling typos noticed along the way.

Signed-off-by: Paul Durrant <paul.durrant@citrix.com>
Reviewed-by: David Vrabel <david.vrabel@citrix.com>
Acked-by: Jan Beulich <jbeulich@suse.com>

errno: declare aliases using XEN_ERRNO()

Otherwise a custom XEN_ERRNO definition will not end up creating appropriately
namespaced constants for the aliases.

Signed-off-by: Andrew Cooper <andrew.cooper3@citrix.com>
Reviewed-by: Doug Goldstein <cardoe@cardoe.com>
Acked-by: Jan Beulich <jbeulich@suse.com>

AMD, maintainers: remove myself from list

I will not be looking at AMD related Xen code now.
So, removing myself.

Signed-off-by: Aravind Gopalakrishnan <aravind.gopalakrishnan@amd.com>
Acked-by: Boris Ostrovsky <boris.ostrovsky@oracle.com>

arm/acpi: Add acpi parameter to enable/disable acpi

Define new command line parameter "acpi" to enable/disable acpi.
This implements the following policy to decide whether ACPI should be
used to boot the system:
- acpi=off: ACPI will not be used to boot the system, even if there is
  no alternative available (e.g., device tree is empty)
- acpi=force: only ACPI will be used to boot the system; if that fails,
  there will be no fallback to alternative methods (such as device tree)
- otherwise, ACPI will be used as a fallback if the device tree turns
  out to lack a platform description; the heuristic to decide this is
  whether /chosen is the only node present at depth 1

Signed-off-by: Shannon Zhao <shannon.zhao@linaro.org>
Reviewed-by: Stefano Stabellini <stefano.stabellini@eu.citrix.com>

arm/fdt: Export device_tree_for_each_node

This function will be used by ACPI booting. Export it so that it can be
used by other files.

Signed-off-by: Shannon Zhao <shannon.zhao@linaro.org>
Reviewed-by: Stefano Stabellini <stefano.stabellini@eu.citrix.com>

arm/acpi: Add a new ACPI initialized function for UART

This adds a new function to initialize UART for ACPI on ARM.

Signed-off-by: Shannon Zhao <shannon.zhao@linaro.org>
Reviewed-by: Stefano Stabellini <stefano.stabellini@eu.citrix.com>

arm/acpi: Parse GTDT to initialize timer

Parse GTDT (Generic Timer Descriptor Table) to initialize timer. Using
the information presented by GTDT to initialize the arch timer (not
memory-mapped).

Signed-off-by: Parth Dixit <parth.dixit@linaro.org>
Signed-off-by: Shannon Zhao <shannon.zhao@linaro.org>
Reviewed-by: Stefano Stabellini <stefano.stabellini@eu.citrix.com>

arm/irq: Add helper function for setting interrupt type

Add a helper function to set edge/level type information for an
interrupt.

Signed-off-by: Parth Dixit <parth.dixit@linaro.org>
Signed-off-by: Shannon Zhao <shannon.zhao@linaro.org>
Acked-by: Stefano Stabellini <stefano.stabellini@eu.citrix.com>

arm/gic: Add ACPI support for GIC preinit

Since ACPI 6.0 defines that GIC Distributor Structure contains the GIC
version filed, it could get GIC version from that. Then call acpi device
initializing function to preinit GIC device.

Signed-off-by: Shannon Zhao <shannon.zhao@linaro.org>
Reviewed-by: Stefano Stabellini <stefano.stabellini@eu.citrix.com>

arm/gic-v3: Add ACPI boot support for GICv3

Like GICv2, ACPI on Xen hypervisor uses MADT table for proper GICv3
initialization. Parse GIC distributor subtable, redistributor subtable
and interrupt subtable.

Signed-off-by: Shannon Zhao <shannon.zhao@linaro.org>
Reviewed-by: Stefano Stabellini <stefano.stabellini@eu.citrix.com>

arm/gic-v2: Add ACPI boot support for GICv2

ACPI on Xen hypervisor uses MADT table for proper GIC initialization.
First get the GIC version from GIC Distributor. Then parse GIC related
subtables, collect CPU interface and distributor addresses and call
driver initialization function (which is hardware abstraction agnostic).
In a similar way, FDT initialize GICv2.

Signed-off-by: Parth Dixit <parth.dixit@linaro.org>
Signed-off-by: Shannon Zhao <shannon.zhao@linaro.org>
Reviewed-by: Stefano Stabellini <stefano.stabellini@eu.citrix.com>

arm/irq: Drop the DT prefix of the irq line type

Make these types generic to DT and ACPI. So they are can be used in ACPI
codes.

Signed-off-by: Shannon Zhao <shannon.zhao@linaro.org>
Reviewed-by: Stefano Stabellini <stefano.stabellini@eu.citrix.com>

arm: Introduce a generic way to use a device from acpi

Add generic way to use device from acpi similar to the way it is
supported in device tree.

Signed-off-by: Parth Dixit <parth.dixit@linaro.org>
Signed-off-by: Shannon Zhao <shannon.zhao@linaro.org>
Acked-by: Stefano Stabellini <stefano.stabellini@eu.citrix.com>

acpi/table: Introduce acpi_table_get_entry_madt to get specified entry

This function could get the specified index entry of MADT table. This
would be useful when it needs to get the contens of the entry.

Signed-off-by: Shannon Zhao <shannon.zhao@linaro.org>
Acked-by: Jan Beulich <jbeulich@suse.com>

arm/acpi: Add ACPI support for SMP initialization

ACPI 5.1 only has two explicit methods to boot up SMP, PSCI and Parking
protocol, but the Parking protocol is only specified for ARMv7 now, so
make PSCI as the only way for the SMP boot protocol before some updates
for the ACPI spec or the Parking protocol spec.

ACPI only supports PSCI 0.2+, since prior to PSCI 0.2 function IDs are
not well-defined.

Signed-off-by: Hanjun Guo <hanjun.guo@linaro.org>
Signed-off-by: Parth Dixit <parth.dixit@linaro.org>
Signed-off-by: Shannon Zhao <shannon.zhao@linaro.org>
Reviewed-by: Stefano Stabellini <stefano.stabellini@eu.citrix.com>

arm/acpi: Parse MADT to map logical cpu to MPIDR and get cpu_possible_map

MADT contains the information for MPIDR which is essential for SMP
initialization, parse the GIC cpu interface structures to get the MPIDR
value and map it to cpu_logical_map(), and add enabled cpu with valid
MPIDR into cpu_possible_map.

Move BAD_MADT_ENTRY to common place, parenthesize its parameters and
drop the pointer cast.

Signed-off-by: Hanjun Guo <hanjun.guo@linaro.org>
Signed-off-by: Tomasz Nowicki <tomasz.nowicki@linaro.org>
Signed-off-by: Naresh Bhat <naresh.bhat@linaro.org>
Signed-off-by: Parth Dixit <parth.dixit@linaro.org>
Signed-off-by: Shannon Zhao <shannon.zhao@linaro.org>
Reviewed-by: Stefano Stabellini <stefano.stabellini@eu.citrix.com>
Acked-by: Jan Beulich <jbeulich@suse.com>

arm/acpi: Parse FADT table and get PSCI flags

There are two flags: PSCI_COMPLIANT and PSCI_USE_HVC. When set, the
former signals to the OS that the hardware is PSCI compliant. The latter
selects the appropriate conduit for PSCI calls by toggling between
Hypervisor Calls (HVC) and Secure Monitor Calls (SMC). FADT table
contains such information, parse FADT to get the flags for furture
usage.

Since STAO table and the GIC version are introduced by ACPI 6.0, we will
check the version and only parse FADT table with version >= 6.0. If
firmware provides ACPI tables with ACPI version less than 6.0, OS will
be messed up with those information, so disable ACPI if we get an FADT
table with version less than 6.0.

Signed-off-by: Hanjun Guo <hanjun.guo@linaro.org>
Signed-off-by: Naresh Bhat <naresh.bhat@linaro.org>
Signed-off-by: Parth Dixit <parth.dixit@linaro.org>
Signed-off-by: Shannon Zhao <shannon.zhao@linaro.org>
Reviewed-by: Stefano Stabellini <stefano.stabellini@eu.citrix.com>

arm/acpi: Move end_boot_allocator after acpi_boot_table_init

To support ACPI NUMA, it needs to make the ACPI initialization done
before boot_end_allocator. Also, x86 does this by the same way.

Signed-off-by: Parth Dixit <parth.dixit@linaro.org>
Signed-off-by: Shannon Zhao <shannon.zhao@linaro.org>
Reviewed-by: Stefano Stabellini <stefano.stabellini@eu.citrix.com>

arm/acpi: Add basic ACPI initialization

acpi_boot_table_init() will be called in start_xen to get the RSDP and
all the table pointers. With this patch, we can get ACPI boot-time
tables from firmware on ARM64.

Signed-off-by: Naresh Bhat <naresh.bhat@linaro.org>
Signed-off-by: Parth Dixit <parth.dixit@linaro.org>
Signed-off-by: Shannon Zhao <shannon.zhao@linaro.org>
Acked-by: Stefano Stabellini <stefano.stabellini@eu.citrix.com>

arm/acpi: Add __acpi_map_table function for ARM

Implement __acpi_map_table function for ARM. Move FIX_ACPI_PAGES to
common place and rename it to NUM_FIXMAP_ACPI_PAGES.

Signed-off-by: Shannon Zhao <shannon.zhao@linaro.org>
Reviewed-by: Stefano Stabellini <stefano.stabellini@eu.citrix.com>
Acked-by: Jan Beulich <jbeulich@suse.com>

errno: introduce EISDIR/EROFS/ENOTEMPTY to the ABI

These POSIX errnos are expected by other areas of the Xen public interface,
specifically public/io/xs_wire.h

Signed-off-by: Andrew Cooper <andrew.cooper3@citrix.com>
Reviewed-by: Doug Goldstein <cardoe@cardoe.com>
Acked-by: Jan Beulich <jbeulich@suse.com>

arm: remove unneeded ifdef CONFIG_ARM in efi-boot.h

Ifdef'ing CONFIG_ARM in xen/arch/arm/efi/efi-boot.h is redundant, remove
the condition and simplify the ifdef's.

Signed-off-by: Stefano Stabellini <stefano.stabellini@eu.citrix.com>
Reviewed-by: Jan Beulich <jbeulich@suse.com>

silence affinity messages on suspend/resume

When taking cpus offline for suspend or bringing them online on resume
again the scheduler might issue debug messages when temporarily
breaking vcpu affinity or restoring the original affinity settings.

The resume message can be removed completely, while the message when
breaking affinity should only be issued if the breakage is permanent.

Suggested-by: Jan Beulich <jbeulich@suse.com>
Signed-off-by: Juergen Gross <jgross@suse.com>
Acked-by: Dario Faggioli <dario.faggioli@citrix.com>

Remus: update email address in MAINTAINERS file

Signed-off-by: Yang Hongyang <imhy.yang@gmail.com>
Cc: Shriram Rajagopalan <rshriram@cs.ubc.ca>
Cc: Wei Liu <wei.liu2@citrix.com>
Cc: Ian Jackson <ian.jackson@eu.citrix.com>
Cc: Ian Campbell <ian.campbell@citrix.com>

travis: add IRC notifications

This will cause failed builds and when the build flips back to success
to be reported to #xentest on FreeNode. The syntax of the message will
be:
<travis-ci> xen-project/xen#BUILDID (BRANCH - REVISION : COMMITTER)
<travis-ci> Change view :
https://github.com/xen-project/xen/compare/RANGE
<travis-ci> Build details :
https://travis-ci.org/xen-project/xen/builds/BUILDID

The blob was generated with the following command:
travis encrypt -r xen-project/xen 'chat.freenode.net#xentest'

The reason it is encrypted is to prevent people that fork the repo to
spam #xentest. This value will only properly decrypt when running within
the xen-project/xen space.

Signed-off-by: Doug Goldstein <cardoe@cardoe.com>
Acked-by: Ian Jackson <ian.jackson@eu.citrix.com>

m4/python: fix checks for Python library support

AC_CHECK_LIB() was running gcc -Llib -lm -lutils conftest.c which on
platforms that do as needed operations by default will result in
underlinking. Instead AC_CHECK_LIB() suggests supplying the extra
libraries necessary in a 5th argument.

Signed-off-by: Doug Goldstein <cardoe@cardoe.com>
Acked-by: Ian Jackson <ian.jackson@eu.citrix.com>

libxc: Expose the MPX cpuid flag to guest

If hardware support memory protect externsion, expose this feature
to guest by default. Users don't have to use a 'cpuid= ' option in
config file to turn it on.

Signed-off-by: Liang Li <liang.z.li@intel.com>
Reviewed-by: Andrew Cooper <andrew.cooper3@citrix.com>

tools/python/xs: fix two comments

These two functions require transaction handle as the first argument.

Reported-by: Sergei Lebedev <sergei.a.lebedev@gmail.com>
Signed-off-by: Wei Liu <wei.liu2@citrix.com>

libxl: fix comment LIBXL__EGC_INIT -> LIBXL_EGC_INIT

Signed-off-by: Wei Liu <wei.liu2@citrix.com>
Cc: Ian Jackson <ian.jackson@eu.citrix.com>

tools/configure: only require bcc/ld86/as86 when needed

bcc/ld86/as86 are necessary when we build ROMBIOS. However if we do not
build it (and are not building qemu-trad), the build requirements are
overly strict and can lead to failures.

Signed-off-by: Doug Goldstein <cardoe@cardoe.com>
Reviewed-by: Konrad Rzeszutek Wilk <konrad.wilk@oracle.com>

libxl: qmp: ensure qmp read buffer is NULL terminated

Coverity rightly points out that qmp->buffer may not be NULL
terminated when passed to strncat.

Make the actual buffer a byte bigger than QMP_RECEIVE_BUFFER_SIZE and
always append a NULL byte.

I suspect that in practice we have not yet seen QMP messages
approaching the buffer size (4K).

Compile tested only.

CID: 1055989

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

xl: NULL terminate buf when reading dom0 /proc/uptime

The contents of /proc/uptime is typically something like "80164.57
640617.58", so the existing 512 byte buffer is more than large enoguh,
so reduce its effective size to 511 bytes and ensure we include a
NULL.

Otherwise Coverity points out that we pass a potentially unterminated
string to strtok. In practice this likely doesn't actually cause
issues (at least on Linux) because the
string should always contain a space so we will stop parsing.

CID: 105590

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

xl: uptime: skip dom0 when calling print_domU_uptime

Dom0 is handled separately (via print_dom0_uptime) and the domU
variant doesn't work for dom0 since libxl_vm_get_start_time() doesn't.

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

docs: Add descriptions of TSC scaling in xl.cfg and tscmode.txt

Signed-off-by: Haozhong Zhang <haozhong.zhang@intel.com>
Reviewed-by: Kevin Tian <kevin.tian@intel.com>

remove unused EXPORT_SYMBOL_GPL

EXPORT_SYMBOL_GPL is unused in the source tree so just remove it. If
something gets imported that needs it, it can be added back then.

Signed-off-by: Doug Goldstein <cardoe@cardoe.com>
Acked-by: Jan Beulich <jbeulich@suse.com>

VT-d: add a command line parameter for posted-interrupts

Enable VT-d Posted-Interrupts and add a command line
parameter for it.

Signed-off-by: Feng Wu <feng.wu@intel.com>
Reviewed-by: Kevin Tian <kevin.tian@intel.com>
Acked-by: Jan Beulich <jbeulich@suse.com>

vmx: VT-d posted-interrupt core logic handling

This is the core logic handling for VT-d posted-interrupts. Basically it
deals with how and when to update posted-interrupts during the following
scenarios:
- vCPU is preempted
- vCPU is slept
- vCPU is blocked

When vCPU is preempted/slept, we update the posted-interrupts during
scheduling by introducing two new architecutral scheduler hooks:
vmx_pi_switch_from() and vmx_pi_switch_to(). When vCPU is blocked, we
introduce a new architectural hook: arch_vcpu_block() to update
posted-interrupts descriptor.

Besides that, before VM-entry, we will make sure the 'NV' filed is set
to 'posted_intr_vector' and the vCPU is not in any blocking lists, which
is needed when vCPU is running in non-root mode. The reason we do this check
is because we change the posted-interrupts descriptor in vcpu_block(),
however, we don't change it back in vcpu_unblock() or when vcpu_block()
directly returns due to event delivery (in fact, we don't need to do it
in the two places, that is why we do it before VM-Entry).

When we handle the lazy context switch for the following two scenarios:
- Preempted by a tasklet, which uses in an idle context.
- the prev vcpu is in offline and no new available vcpus in run queue.
We don't change the 'SN' bit in posted-interrupt descriptor, this
may incur spurious PI notification events, but since PI notification
event is only sent when 'ON' is clear, and once the PI notificatoin
is sent, ON is set by hardware, hence no more notification events
before 'ON' is clear. Besides that, spurious PI notification events are
going to happen from time to time in Xen hypervisor, such as, when
guests trap to Xen and PI notification event happens, there is
nothing Xen actually needs to do about it, the interrupts will be
delivered to guest atht the next time we do a VMENTRY.

Suggested-by: Yang Zhang <yang.z.zhang@intel.com>
Suggested-by: Dario Faggioli <dario.faggioli@citrix.com>
Suggested-by: George Dunlap <george.dunlap@citrix.com>
Suggested-by: Jan Beulich <jbeulich@suse.com>
Signed-off-by: Feng Wu <feng.wu@intel.com>
Reviewed-by: Jan Beulich <jbeulich@suse.com>
Reviewed-by: Dario Faggioli <dario.faggioli@citrix.com>
Acked-by: Kevin Tian <kevin.tian@intel.com>
Reviewed-by: George Dunlap <george.dunlap@citrix.com>

Merge branch 'staging' of xenbits.xen.org:/home/xen/git/xen into staging

vmx: add VMX RDTSC(P) scaling support

Signed-off-by: Haozhong Zhang <haozhong.zhang@intel.com>
Acked-by: Jan Beulich <jbeulich@suse.com>
Acked-by: Kevin Tian <kevin.tian@intel.com>

xl: close nullfd after dup2'ing it to stdin

We assert that nullfd if not std{in,out,err} since that would result
in closing one of the just dup2'd fds. For this to happen
std{in,out,err} would have needed to be closed, at which point all
sorts of other things could go wrong.

CID: 1130519

It was previously hypothesised[0] that fixing 1130516 would solve this
too, but that appears to not have been the case.

Compile tested only.

[0] http://lists.xenproject.org/archives/html/xen-devel/2013-11/msg02931.html

Signed-off-by: Ian Campbell <ian.campbell@citrix.com>
Cc: andrew.cooper3@citrix.com
Acked-by: Ian Jackson <ian.jackson@eu.citrix.com>

x86/hvm: move saving/loading vcpu's TSC to common code

Both VMX and SVM save/load vcpu's TSC when saving/loading vcpu's
context, so this patch moves saving/loading vcpu's TSC to the common
functions hvm_[save|load]_cpu_ctxt().

Signed-off-by: Haozhong Zhang <haozhong.zhang@intel.com>
Acked-by: Jan Beulich <jbeulich@suse.com>
Acked-by: Kevin Tian <kevin.tian@intel.com>
Reviewed-by: Boris Ostrovsky <boris.ostrovsky@oracle.com>

x86/hvm: replace architecture TSC scaling by a common function

This patch implements a common function hvm_scale_tsc() to scale TSC by
using TSC scaling information collected by architecture code.

Signed-off-by: Haozhong Zhang <haozhong.zhang@intel.com>
Acked-by: Boris Ostrovsky <boris.ostrovsky@oracle.com> for SVM bits
Reviewed-by: Jan Beulich <jbeulich@suse.com>

tools/xenalyze: Actually handle case where number of ipi vectors exceeds static max

find_vec() is supposed to find the vector in the list if it exists,
choose an empty slot if it doesn't exist, and return null if all slots
are full.

However, coverity noticed that although the callers of find_vec() handle
the last condition, find_vec() itself didn't.

Check to see if we actually found an empty slot before attempting to
initialize it.

CID 1306864

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

tools/xenalyze: Fix check for error return value

fdopen returns NULL on failure, not a negative integer.

CID 1306863
CID 1306858

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

tools/xenalyze: Avoid redundant check

Coverity notices that if !head is true, that !N can never be true.

Don't bother checking N, since we know it has to be at least one.

CID 1354243

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

tools/xenalyze: Close symbol_file after reading it

...to avoid leaking the FD and associated memory.

CID 1306872

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

loadpolicy: only close polFd if it is valid

It can be -1 at this point.

CID 1055562
Signed-off-by: Ian Campbell <ian.campbell@citrix.com>
Acked-by: Wei Liu <wei.liu2@citrix.com>

MAINTAINERS: remove myself from ARM (incl DT), seabios, tools and The Rest

Signed-off-by: Ian Campbell <ian.campbell@citrix.com>
Acked-by: Wei Liu <wei.liu2@citrix.com>

build: consolidate CONFIG_HAS_VIDEO and CONFIG_VIDEO

No real advantage to keeping these separate. The use case of this from
Linux is when the platform or target board has support for something but
the user wants to be given the option to disable it.

Signed-off-by: Doug Goldstein <cardoe@cardoe.com>
Acked-by: Jan Beulich <jbeulich@suse.com>