qapi: introduce new cmd option "allow-preconfig"

New option will be used to allow commands, which are prepared/need
to run, during preconfig state. Other commands that should be able
to run in preconfig state, should be amended to not expect machine
in initialized state or deal with it.

For compatibility reasons, commands that don't use new flag
'allow-preconfig' explicitly are not permitted to run in
preconfig state but allowed in all other states like they used
to be.

Within this patch allow following commands in preconfig state:
   qmp_capabilities
   query-qmp-schema
   query-commands
   query-command-line-options
   query-status
   exit-preconfig
to allow qmp connection, basic introspection and moving to the next
state.

PS:
set-numa-node and query-hotpluggable-cpus will be enabled later in
a separate patches.

Signed-off-by: Igor Mammedov <imammedo@redhat.com>
Message-Id: <1526057503-39287-1-git-send-email-imammedo@redhat.com>
Reviewed-by: Eric Blake <eblake@redhat.com>
[ehabkost: Changed "since 2.13" to "since 3.0"]
Signed-off-by: Eduardo Habkost <ehabkost@redhat.com>

hmp: disable monitor in preconfig state

Ban it for now, if someone would need it to work early,
one would have to implement checks if HMP command is valid
at preconfig state.

Signed-off-by: Igor Mammedov <imammedo@redhat.com>
Reviewed-by: Eric Blake <eblake@redhat.com>
Message-Id: <1525423069-61903-5-git-send-email-imammedo@redhat.com>
Signed-off-by: Eduardo Habkost <ehabkost@redhat.com>

qapi: introduce preconfig runstate

New preconfig runstate will be used in follow up patches
related to introducing --preconfig CLI option and is
intended to replace prelaunch runstate from QEMU start
up to machine_init callback.

Signed-off-by: Igor Mammedov <imammedo@redhat.com>
Message-Id: <1525423069-61903-4-git-send-email-imammedo@redhat.com>
Reviewed-by: Eric Blake <eblake@redhat.com>
[ehabkost: Changed "since 2.13" to "since 3.0"]
Signed-off-by: Eduardo Habkost <ehabkost@redhat.com>

numa: split out NumaOptions parsing into set_numa_options()

it will allow to reuse set_numa_options() for parsing
configuration commands received via QMP interface

Signed-off-by: Igor Mammedov <imammedo@redhat.com>
Message-Id: <1525423069-61903-3-git-send-email-imammedo@redhat.com>
Signed-off-by: Eduardo Habkost <ehabkost@redhat.com>

numa: postpone options post-processing till machine_run_board_init()

in preparation for numa options to being handled via QMP before
machine_run_board_init(), move final numa configuration checks
and processing to machine_run_board_init() so it could take into
account both CLI (via parse_numa_opts()) and QMP input

Signed-off-by: Igor Mammedov <imammedo@redhat.com>
Reviewed-by: Eduardo Habkost <ehabkost@redhat.com>
Message-Id: <1525423069-61903-2-git-send-email-imammedo@redhat.com>
Signed-off-by: Eduardo Habkost <ehabkost@redhat.com>

numa: clarify error message when node index is out of range in -numa dist, ...

When using following CLI:
  -numa dist,src=128,dst=1,val=20
user gets a rather confusing error message:
   "Invalid node 128, max possible could be 128"

Where 128 is number of nodes that QEMU supports (MAX_NODES),
while src/dst is an index up to that limit, so it should be
MAX_NODES - 1 in error message.
Make error message to explicitly state valid range for node
index to be more clear.

Signed-off-by: Igor Mammedov <imammedo@redhat.com>
Message-Id: <1526483174-169008-1-git-send-email-imammedo@redhat.com>
Reviewed-by: Eric Blake <eblake@redhat.com>
Signed-off-by: Eduardo Habkost <ehabkost@redhat.com>

Merge remote-tracking branch 'remotes/edgar/tags/edgar/xilinx-next-2018-05-29-v1.for-upstream' into staging

Tag edgar/xilinx-next-2018-05-29-v1.for-upstream

# gpg: Signature made Tue 29 May 2018 09:58:30 BST
# gpg:                using RSA key 29C596780F6BCA83
# gpg: Good signature from "Edgar E. Iglesias (Xilinx key) <edgar.iglesias@xilinx.com>"
# gpg:                 aka "Edgar E. Iglesias <edgar.iglesias@gmail.com>"
# Primary key fingerprint: AC44 FEDC 14F7 F1EB EDBF  4151 29C5 9678 0F6B CA83

* remotes/edgar/tags/edgar/xilinx-next-2018-05-29-v1.for-upstream: (38 commits)
  target-microblaze: Consolidate MMU enabled checks
  target-microblaze: cpu_mmu_index: Fixup indentation
  target-microblaze: Use tcg_gen_movcond in eval_cond_jmp
  target-microblaze: Convert env_btarget to i64
  target-microblaze: Remove argument b in eval_cc()
  target-microblaze: Use table based condition-codes conversion
  target-microblaze: mmu: Cleanup debug log messages
  target-microblaze: Simplify address computation using tcg_gen_addi_i32()
  target-microblaze: Allow address sizes between 32 and 64 bits
  target-microblaze: Add support for extended access to TLBLO
  target-microblaze: dec_msr: Plug a temp leak
  target-microblaze: mmu: Add a configurable output address mask
  target-microblaze: mmu: Prepare for 64-bit addresses
  target-microblaze: mmu: Remove unused register state
  target-microblaze: mmu: Add R_TBLX_MISS macros
  target-microblaze: Implement MFSE EAR
  target-microblaze: Add Extended Addressing
  target-microblaze: Setup for 64bit addressing
  target-microblaze: Make special registers 64-bit
  target-microblaze: dec_msr: Fix MTS to FSR
  ...

Signed-off-by: Peter Maydell <peter.maydell@linaro.org>

qemu-doc.texi: Rename references to 2.13 to 3.0

Update references to 2.13 to read 3.0, since that's the
number we're using for the next release.

Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
Reviewed-by: Cornelia Huck <cohuck@redhat.com>
Reviewed-by: Thomas Huth <thuth@redhat.com>
Message-id: 20180522104000.9044-6-peter.maydell@linaro.org

ppc: Rename 2.13 machines to 3.0

Rename the 2.13 machines to match the number we're going to
use for the next release.

Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
Reviewed-by: Cornelia Huck <cohuck@redhat.com>
Reviewed-by: Thomas Huth <thuth@redhat.com>
Reviewed-by: Greg Kurz <groug@kaod.org>
Message-id: 20180522104000.9044-5-peter.maydell@linaro.org

hw/s390x: Rename 2.13 machines to 3.0

Rename the 2.13 machines to match the number we're going to
use for the next release.

Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
Reviewed-by: Cornelia Huck <cohuck@redhat.com>
Reviewed-by: Thomas Huth <thuth@redhat.com>
Message-id: 20180522104000.9044-4-peter.maydell@linaro.org

hw/i386: Rename 2.13 machine types to 3.0

Rename the 2.13 machine types to match what we're going to
use as our next release number.

Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
Reviewed-by: Cornelia Huck <cohuck@redhat.com>
Reviewed-by: Thomas Huth <thuth@redhat.com>
Reviewed-by: Eduardo Habkost <ehabkost@redhat.com>
Message-id: 20180522104000.9044-3-peter.maydell@linaro.org

qapi: Change "since 2.13" annotations to "since 3.0"

We're going to make the next release be 3.0, not 2.13; change
the annotations in our json appropriately.

Changes produced with
  sed -i -e 's/2\.13/3.0/g' qapi/*.json

Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
Reviewed-by: Cornelia Huck <cohuck@redhat.com>
Reviewed-by: Thomas Huth <thuth@redhat.com>
Message-id: 20180522104000.9044-2-peter.maydell@linaro.org

Merge remote-tracking branch 'remotes/stefanberger/tags/pull-tpm-2018-05-23-4' into staging

Merge tpm 2018/05/23 v4

# gpg: Signature made Sat 26 May 2018 03:52:12 BST
# gpg:                using RSA key 75AD65802A0B4211
# gpg: Good signature from "Stefan Berger <stefanb@linux.vnet.ibm.com>"
# gpg: WARNING: This key is not certified with a trusted signature!
# gpg:          There is no indication that the signature belongs to the owner.
# Primary key fingerprint: B818 B9CA DF90 89C2 D5CE  C66B 75AD 6580 2A0B 4211

* remotes/stefanberger/tags/pull-tpm-2018-05-23-4:
  test: Add test cases that use the external swtpm with CRB interface
  docs: tpm: add VM save/restore example and troubleshooting guide
  tpm: extend TPM TIS with state migration support
  tpm: extend TPM emulator with state migration support

Signed-off-by: Peter Maydell <peter.maydell@linaro.org>

target-microblaze: Consolidate MMU enabled checks

Consolidate MMU enabled checks to cpu_mmu_index().
No functional changes.

Suggested-by: Richard Henderson <richard.henderson@linaro.org>
Reviewed-by: Alistair Francis <alistair.francis@wdc.com>
Reviewed-by: Richard Henderson <richard.henderson@linaro.org>
Signed-off-by: Edgar E. Iglesias <edgar.iglesias@xilinx.com>

target-microblaze: cpu_mmu_index: Fixup indentation

Fixup the indentation of cpu_mmu_index in preparation for
future edits.
No functional changes.

Reviewed-by: Alistair Francis <alistair.francis@wdc.com>
Reviewed-by: Richard Henderson <richard.henderson@linaro.org>
Signed-off-by: Edgar E. Iglesias <edgar.iglesias@xilinx.com>

target-microblaze: Use tcg_gen_movcond in eval_cond_jmp

Cleanup eval_cond_jmp to use tcg_gen_movcond_i64().
No functional change.

Suggested-by: Richard Henderson <richard.henderson@linaro.org>
Reviewed-by: Philippe Mathieu-Daudé <f4bug@amsat.org>
Reviewed-by: Richard Henderson <richard.henderson@linaro.org>
Signed-off-by: Edgar E. Iglesias <edgar.iglesias@xilinx.com>

target-microblaze: Convert env_btarget to i64

Convert env_btarget to i64.
No functional change.

Reviewed-by: Philippe Mathieu-Daudé <f4bug@amsat.org>
Reviewed-by: Richard Henderson <richard.henderson@linaro.org>
Signed-off-by: Edgar E. Iglesias <edgar.iglesias@xilinx.com>

target-microblaze: Remove argument b in eval_cc()

Remove argument b in eval_cc() as it is always set to zero.
No functional change.

Reviewed-by: Richard Henderson <richard.henderson@linaro.org>
Signed-off-by: Edgar E. Iglesias <edgar.iglesias@xilinx.com>

target-microblaze: Use table based condition-codes conversion

Use a table based conversion to map condition-codes between
MicroBlaze ISA encoding and TCG.
No functional change.

Reviewed-by: Richard Henderson <richard.henderson@linaro.org>
Signed-off-by: Edgar E. Iglesias <edgar.iglesias@xilinx.com>

target-microblaze: mmu: Cleanup debug log messages

Cleanup debug log messages:
* Avoid long 80+ character lines.
* Remove D() macro and use qemu_log_mask.
* Remove logs that are not very useful

Suggested-by: Alistair Francis <alistair.francis@wdc.com>
Reviewed-by: Richard Henderson <richard.henderson@linaro.org>
Reviewed-by: Alistair Francis <alistair.francis@wdc.com>
Signed-off-by: Edgar E. Iglesias <edgar.iglesias@xilinx.com>

target-microblaze: Simplify address computation using tcg_gen_addi_i32()

Simplify address computation using tcg_gen_addi_i32().
tcg_gen_addi_i32() already optimizes the case when the
immediate is zero.

No functional change.

Suggested-by: Richard Henderson <richard.henderson@linaro.org>
Reviewed-by: Richard Henderson <richard.henderson@linaro.org>
Reviewed-by: Philippe Mathieu-Daudé <f4bug@amsat.org>
Signed-off-by: Edgar E. Iglesias <edgar.iglesias@xilinx.com>

target-microblaze: Allow address sizes between 32 and 64 bits

Allow address sizes between 32 and 64 bits.

Reviewed-by: Alistair Francis <alistair.francis@wdc.com>
Reviewed-by: Richard Henderson <richard.henderson@linaro.org>
Signed-off-by: Edgar E. Iglesias <edgar.iglesias@xilinx.com>

target-microblaze: Add support for extended access to TLBLO

Add support for extended access to TLBLO's upper 32 bits.

Reviewed-by: Alistair Francis <alistair.francis@wdc.com>
Signed-off-by: Edgar E. Iglesias <edgar.iglesias@xilinx.com>

target-microblaze: dec_msr: Plug a temp leak

Plug a temp leak.

Reported-by: Richard Henderson <richard.henderson@linaro.org>
Reviewed-by: Philippe Mathieu-Daudé <f4bug@amsat.org>
Signed-off-by: Edgar E. Iglesias <edgar.iglesias@xilinx.com>

target-microblaze: mmu: Add a configurable output address mask

Add a configurable output address mask, used to mimic the
configurable physical address bit width.

Reviewed-by: Alistair Francis <alistair.francis@wdc.com>
Reviewed-by: Richard Henderson <richard.henderson@linaro.org>
Signed-off-by: Edgar E. Iglesias <edgar.iglesias@xilinx.com>

target-microblaze: mmu: Prepare for 64-bit addresses

Prepare for 64-bit addresses.
This makes no functional difference as the upper parts of
the 64-bit addresses are not yet reachable.

Reviewed-by: Richard Henderson <richard.henderson@linaro.org>
Signed-off-by: Edgar E. Iglesias <edgar.iglesias@xilinx.com>

target-microblaze: mmu: Remove unused register state

Add explicit handling for MMU_R_TLBX and log accesses to
invalid MMU registers. We can now remove the state for
all regs but PID, ZPR and TLBX (0 - 2).

Reviewed-by: Richard Henderson <richard.henderson@linaro.org>
Reviewed-by: Philippe Mathieu-Daudé <f4bug@amsat.org>
Signed-off-by: Edgar E. Iglesias <edgar.iglesias@xilinx.com>

target-microblaze: mmu: Add R_TBLX_MISS macros

Add a R_TBLX_MISS MASK and SHIFT macros.

Reviewed-by: Alistair Francis <alistair.francis@wdc.com>
Reviewed-by: Richard Henderson <richard.henderson@linaro.org>
Signed-off-by: Edgar E. Iglesias <edgar.iglesias@xilinx.com>

target-microblaze: Implement MFSE EAR

Implement MFSE EAR to enable access to the upper part of EAR.

Reviewed-by: Richard Henderson <richard.henderson@linaro.org>
Signed-off-by: Edgar E. Iglesias <edgar.iglesias@xilinx.com>

target-microblaze: Add Extended Addressing

Add support for Extended Addressing. Load/stores with EA
enabled concatenate two 32bit registers to form an extended
address.

We don't allow users to enable address sizes larger than
32 bits quite yet though. Once the MMU support is in, we'll
turn it on.

Reviewed-by: Richard Henderson <richard.henderson@linaro.org>
Signed-off-by: Edgar E. Iglesias <edgar.iglesias@xilinx.com>

target-microblaze: Setup for 64bit addressing

Setup MicroBlaze builds for 64bit addressing.
No functional change since the translator does not yet
emit 64bit addresses.

Reviewed-by: Richard Henderson <richard.henderson@linaro.org>
Signed-off-by: Edgar E. Iglesias <edgar.iglesias@xilinx.com>

target-microblaze: Make special registers 64-bit

Extend special registers to 64-bits. This is in preparation for
MFSE/MTSE, moves to and from extended special registers.

Reviewed-by: Richard Henderson <richard.henderson@linaro.org>
Signed-off-by: Edgar E. Iglesias <edgar.iglesias@xilinx.com>

target-microblaze: dec_msr: Fix MTS to FSR

Fix moves to FSR. Not only bit 31 is accessible.

Reviewed-by: Alistair Francis <alistair.francis@wdc.com>
Reviewed-by: Richard Henderson <richard.henderson@linaro.org>
Signed-off-by: Edgar E. Iglesias <edgar.iglesias@xilinx.com>

target-microblaze: dec_msr: Reuse more code when reg-decoding

Reuse more code when decoding register numbers.

No functional changes.

Reviewed-by: Alistair Francis <alistair.francis@wdc.com>
Reviewed-by: Richard Henderson <richard.henderson@linaro.org>
Signed-off-by: Edgar E. Iglesias <edgar.iglesias@xilinx.com>

target-microblaze: dec_msr: Use bool and extract32

Use bool and extract32 to represent the to, clr and
clrset flags.

No functional change.

Reviewed-by: Alistair Francis <alistair.francis@wdc.com>
Reviewed-by: Richard Henderson <richard.henderson@linaro.org>
Signed-off-by: Edgar E. Iglesias <edgar.iglesias@xilinx.com>

target-microblaze: Break out trap_illegal()

Break out trap_illegal() to handle illegal operation traps.
We now generally stop translation of the current insn if
it's not valid.

Reviewed-by: Richard Henderson <richard.henderson@linaro.org>
Reviewed-by: Philippe Mathieu-Daudé <f4bug@amsat.org>
Signed-off-by: Edgar E. Iglesias <edgar.iglesias@xilinx.com>

target-microblaze: Break out trap_userspace()

Break out trap_userspace() to avoid open coding it everywhere.
For privileged insns, we now always stop translation of the
current insn for cores without exceptions.

Reviewed-by: Richard Henderson <richard.henderson@linaro.org>
Reviewed-by: Philippe Mathieu-Daudé <f4bug@amsat.org>
Signed-off-by: Edgar E. Iglesias <edgar.iglesias@xilinx.com>

target-microblaze: Name special registers we support

Name special registers we support.

Reviewed-by: Alistair Francis <alistair.francis@wdc.com>
Signed-off-by: Edgar E. Iglesias <edgar.iglesias@xilinx.com>

target-microblaze: Use TCGv for load/store addresses

Use TCGv for load/store addresses, allowing for future
computation of 64-bit load/store address.

No functional change.

Acked-by: Alistair Francis <alistair.francis@wdc.com>
Reviewed-by: Richard Henderson <richard.henderson@linaro.org>
Signed-off-by: Edgar E. Iglesias <edgar.iglesias@xilinx.com>

target-microblaze: Remove pointer indirection for ld/st addresses

Reviewed-by: Alistair Francis <alistair.francis@wdc.com>
Reviewed-by: Richard Henderson <richard.henderson@linaro.org>
Signed-off-by: Edgar E. Iglesias <edgar.iglesias@xilinx.com>

target-microblaze: Make compute_ldst_addr always use a temp

Make compute_ldst_addr always use a temp. This simplifies
the code a bit in preparation for adding support for
64bit addresses.

No functional change.

Reviewed-by: Richard Henderson <richard.henderson@linaro.org>
Reviewed-by: Philippe Mathieu-Daudé <f4bug@amsat.org>
Signed-off-by: Edgar E. Iglesias <edgar.iglesias@xilinx.com>

target-microblaze: Bypass MMU with MMU_NOMMU_IDX

Bypass MMU translation when mmu-index MMU_NOMMU_IDX is used.

Reviewed-by: Alistair Francis <alistair.francis@wdc.com>
Reviewed-by: Richard Henderson <richard.henderson@linaro.org>
Signed-off-by: Edgar E. Iglesias <edgar.iglesias@xilinx.com>

target-microblaze: Conditionalize setting of PVR11_USE_MMU

Conditionalize setting of PVR11_USE_MMU on the use_mmu
CPU property, otherwise we may incorrectly advertise an
MMU via PVR when the core in fact has none.

Reviewed-by: Alistair Francis <alistair.francis@wdc.com>
Reviewed-by: Richard Henderson <richard.henderson@linaro.org>
Signed-off-by: Edgar E. Iglesias <edgar.iglesias@xilinx.com>

target-microblaze: Remove USE_MMU PVR checks

We already have a CPU property to control if a core has
an MMU or not. Remove USE_MMU PVR checks in favor of
looking at the property.

Reviewed-by: Alistair Francis <alistair.francis@wdc.com>
Reviewed-by: Richard Henderson <richard.henderson@linaro.org>
Signed-off-by: Edgar E. Iglesias <edgar.iglesias@xilinx.com>

target-microblaze: Tighten up TCGv_i32 vs TCGv type usage

Tighten up TCGv_i32 vs TCGv type usage. Avoid using TCGv when
TCGv_i32 should be used.

This is in preparation for adding 64bit addressing support.
No functional change.

Reviewed-by: Alistair Francis <alistair.francis@wdc.com>
Reviewed-by: Richard Henderson <richard.henderson@linaro.org>
Signed-off-by: Edgar E. Iglesias <edgar.iglesias@xilinx.com>

target-microblaze: Correct the PVR array size

Correct the PVR array size, there are 13 PVR registers.

Reviewed-by: Alistair Francis <alistair.francis@wdc.com>
Signed-off-by: Edgar E. Iglesias <edgar.iglesias@xilinx.com>

target-microblaze: Correct special register array sizes

Correct special register array sizes.

Reviewed-by: Richard Henderson <richard.henderson@linaro.org>
Signed-off-by: Edgar E. Iglesias <edgar.iglesias@xilinx.com>

target-microblaze: Fallback to our latest CPU version

Today, when running QEMU in linux-user or with boards that don't
select a specific CPU version, we treat it as an invalid version
and log a message.

Instead, if no specific version was selected, fallback to our
latest CPU version.

Reviewed-by: Alistair Francis <alistair.francis@wdc.com>
Signed-off-by: Edgar E. Iglesias <edgar.iglesias@xilinx.com>

target-microblaze: compute_ldst_addr: Use bool instead of int

Use bool instead of int to represent flags.
No functional change.

Reviewed-by: Alistair Francis <alistair.francis@wdc.com>
Reviewed-by: Richard Henderson <richard.henderson@linaro.org>
Signed-off-by: Edgar E. Iglesias <edgar.iglesias@xilinx.com>

target-microblaze: dec_store: Use bool instead of unsigned int

Use bool instead of unsigned int to represent flags.
Also, use extract32 instead of open coding the bit extract.

No functional change.

Reviewed-by: Alistair Francis <alistair.francis@wdc.com>
Reviewed-by: Richard Henderson <richard.henderson@linaro.org>
Signed-off-by: Edgar E. Iglesias <edgar.iglesias@xilinx.com>

target-microblaze: dec_load: Use bool instead of unsigned int

Use bool instead of unsigned int to represent flags.
No functional change.

Reviewed-by: Alistair Francis <alistair.francis@wdc.com>
Reviewed-by: Richard Henderson <richard.henderson@linaro.org>
Signed-off-by: Edgar E. Iglesias <edgar.iglesias@xilinx.com>

test: Add test cases that use the external swtpm with CRB interface

Add a test program for testing the CRB with the external swtpm.

The 1st test case extends a PCR and reads back the value and compares
it against an expected return packet.

The 2nd test case repeats the 1st test case and then migrates the
external swtpm's state along with the VM state to a destination
QEMU and swtpm and checks that the PCR has the expected value now.

The test cases require 'swtpm' to be installed on the system and
in the PATH and 'swtpm' must support the --tpm2 option. If this is
not the case, the test will be skipped.

Signed-off-by: Stefan Berger <stefanb@linux.vnet.ibm.com>
Reviewed-by: Marc-André Lureau <marcandre.lureau@redhat.com>

Merge remote-tracking branch 'remotes/juanquintela/tags/migration/20180525' into staging

migration/next for 20180525

# gpg: Signature made Fri 25 May 2018 14:30:47 BST
# gpg:                using RSA key F487EF185872D723
# gpg: Good signature from "Juan Quintela <quintela@redhat.com>"
# gpg:                 aka "Juan Quintela <quintela@trasno.org>"
# Primary key fingerprint: 1899 FF8E DEBF 58CC EE03  4B82 F487 EF18 5872 D723

* remotes/juanquintela/tags/migration/20180525:
  migration: use g_free for ram load bitmap
  migration: fix exec/fd migrations

Signed-off-by: Peter Maydell <peter.maydell@linaro.org>

migration: use g_free for ram load bitmap

Buffers allocated with bitmap_new() should be freed with g_free().

Both reported by Coverity:

*** CID 1391300:  API usage errors  (ALLOC_FREE_MISMATCH)
/migration/ram.c: 3517 in ram_dirty_bitmap_reload()
3511          * the last one to sync, we need to notify the main send thread.
3512          */
3513         ram_dirty_bitmap_reload_notify(s);
3514
3515         ret = 0;
3516     out:
>>>     CID 1391300:  API usage errors  (ALLOC_FREE_MISMATCH)
>>>     Calling "free" frees "le_bitmap" using "free" but it should have been freed using "g_free".
3517         free(le_bitmap);
3518         return ret;
3519     }
3520
3521     static int ram_resume_prepare(MigrationState *s, void *opaque)
3522     {

*** CID 1391292:  API usage errors  (ALLOC_FREE_MISMATCH)
/migration/ram.c: 249 in ramblock_recv_bitmap_send()
243          * Mark as an end, in case the middle part is screwed up due to
244          * some "misterious" reason.
245          */
246         qemu_put_be64(file, RAMBLOCK_RECV_BITMAP_ENDING);
247         qemu_fflush(file);
248
>>>     CID 1391292:  API usage errors  (ALLOC_FREE_MISMATCH)
>>>     Calling "free" frees "le_bitmap" using "free" but it should have been freed using "g_free".
249         free(le_bitmap);
250
251         if (qemu_file_get_error(file)) {
252             return qemu_file_get_error(file);
253         }
254

Signed-off-by: Peter Xu <peterx@redhat.com>
Reviewed-by: Juan Quintela <quintela@redhat.com>
Message-Id: <20180525015042.31778-1-peterx@redhat.com>
Signed-off-by: Juan Quintela <quintela@redhat.com>

migration: fix exec/fd migrations

Commit:

commit 36c2f8be2c4eb0003ac77a14910842b7ddd7337e
Author: Juan Quintela <quintela@redhat.com>
Date:   Wed Mar 7 08:40:52 2018 +0100

    migration: Delay start of migration main routines

Missed tcp and fd transports.  This fix its.

Reported-by: Kevin Wolf <kwolf@redhat.com>
Signed-off-by: Juan Quintela <quintela@redhat.com>
Tested-by: Kevin Wolf <kwolf@redhat.com>
Message-Id: <20180523091411.1073-1-quintela@redhat.com>
Signed-off-by: Juan Quintela <quintela@redhat.com>

Merge remote-tracking branch 'remotes/vivier2/tags/linux-user-for-2.13-pull-request' into staging

This pull request includes:
- fixes for some comments
- netlink update and fix
- rework/cleanup fo socket.h,
  including fixes for SPARC part.

# gpg: Signature made Fri 25 May 2018 09:16:21 BST
# gpg:                using RSA key F30C38BD3F2FBE3C
# gpg: Good signature from "Laurent Vivier <lvivier@redhat.com>"
# gpg:                 aka "Laurent Vivier <laurent@vivier.eu>"
# gpg:                 aka "Laurent Vivier (Red Hat) <lvivier@redhat.com>"
# Primary key fingerprint: CD2F 75DD C8E3 A4DC 2E4F  5173 F30C 38BD 3F2F BE3C

* remotes/vivier2/tags/linux-user-for-2.13-pull-request:
  gdbstub: Clarify what gdb_handlesig() is doing
  linux-user: define TARGET_SO_REUSEPORT
  linux-user: copy sparc/sockbits.h definitions from linux
  linux-user: update ARCH_HAS_SOCKET_TYPES use
  linux-user: move ppc socket.h definitions to ppc/sockbits.h
  linux-user: move socket.h generic definitions to generic/sockbits.h
  linux-user: move sparc/sparc64 socket.h definitions to sparc/sockbits.h
  linux-user: move alpha socket.h definitions to alpha/sockbits.h
  linux-user: move mips socket.h definitions to mips/sockbits.h
  linux-user: Fix payload size logic in host_to_target_cmsg()
  linux-user: update comments to point to tcg_exec_init()
  linux-user: update netlink emulation
  linux-user: Assert on bad type in thunk_type_align() and thunk_type_size()

Signed-off-by: Peter Maydell <peter.maydell@linaro.org>

gdbstub: Clarify what gdb_handlesig() is doing

gdb_handlesig()'s behaviour is not entirely obvious at first
glance. Add a doc comment for it, and also add a comment
explaining why it's ok for gdb_do_syscallv() to ignore
gdb_handlesig()'s return value. (Coverity complains about
this: CID 1390850.)

Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
Message-Id: <20180515181958.25837-1-peter.maydell@linaro.org>
Signed-off-by: Laurent Vivier <laurent@vivier.eu>

linux-user: define TARGET_SO_REUSEPORT

Signed-off-by: Laurent Vivier <laurent@vivier.eu>
Reviewed-by: Philippe Mathieu-Daudé <f4bug@amsat.org>
Message-Id: <20180519092956.15134-9-laurent@vivier.eu>

linux-user: copy sparc/sockbits.h definitions from linux

Values defined for sparc are not correct.
Copy the content of "arch/sparc/include/uapi/asm/socket.h"
to fix them.

Signed-off-by: Laurent Vivier <laurent@vivier.eu>
Reviewed-by: Peter Maydell <peter.maydell@linaro.org>
Message-Id: <20180519092956.15134-8-laurent@vivier.eu>

linux-user: update ARCH_HAS_SOCKET_TYPES use

to be like in the kernel and rename it TARGET_ARCH_HAS_SOCKET_TYPES

Signed-off-by: Laurent Vivier <laurent@vivier.eu>
Reviewed-by: Peter Maydell <peter.maydell@linaro.org>
Message-Id: <20180519092956.15134-7-laurent@vivier.eu>

linux-user: move ppc socket.h definitions to ppc/sockbits.h

Change conditional #ifdef part by #undef of the symbols
redefined for PPC relative to generic/socket.h

Signed-off-by: Laurent Vivier <laurent@vivier.eu>
Reviewed-by: Peter Maydell <peter.maydell@linaro.org>
Message-Id: <20180519092956.15134-6-laurent@vivier.eu>

linux-user: move socket.h generic definitions to generic/sockbits.h

and include the file from architectures without specific definitions

Signed-off-by: Laurent Vivier <laurent@vivier.eu>
Reviewed-by: Peter Maydell <peter.maydell@linaro.org>
Message-Id: <20180519092956.15134-5-laurent@vivier.eu>

linux-user: move sparc/sparc64 socket.h definitions to sparc/sockbits.h

No code change.

Signed-off-by: Laurent Vivier <laurent@vivier.eu>
Reviewed-by: Peter Maydell <peter.maydell@linaro.org>
Reviewed-by: Philippe Mathieu-Daudé <f4bug@amsat.org>
Message-Id: <20180519092956.15134-4-laurent@vivier.eu>

linux-user: move alpha socket.h definitions to alpha/sockbits.h

No code change.

Signed-off-by: Laurent Vivier <laurent@vivier.eu>
Reviewed-by: Peter Maydell <peter.maydell@linaro.org>
Reviewed-by: Philippe Mathieu-Daudé <f4bug@amsat.org>
Message-Id: <20180519092956.15134-3-laurent@vivier.eu>

linux-user: move mips socket.h definitions to mips/sockbits.h

No code change.

Signed-off-by: Laurent Vivier <laurent@vivier.eu>
Reviewed-by: Peter Maydell <peter.maydell@linaro.org>
Reviewed-by: Philippe Mathieu-Daudé <f4bug@amsat.org>
Message-Id: <20180519092956.15134-2-laurent@vivier.eu>

linux-user: Fix payload size logic in host_to_target_cmsg()

Coverity points out that there's a missing break in the switch in
host_to_target_cmsg() where we update tgt_len for
cmsg_level/cmsg_type combinations which require a different length
for host and target (CID 1385425).  To avoid duplicating the default
case (target length same as host) in both switches, set that before
the switch so that only the cases which want to override it need any
code.

This fixes a bug where we would have used the wrong length
for SOL_SOCKET/SO_TIMESTAMP messages where the target and
host have differently sized 'struct timeval' (ie one is 32
bit and the other is 64 bit).

Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
Reviewed-by: Laurent Vivier <laurent@vivier.eu>
Message-Id: <20180518184715.29833-1-peter.maydell@linaro.org>
Signed-off-by: Laurent Vivier <laurent@vivier.eu>

linux-user: update comments to point to tcg_exec_init()

cpu_init() was replaced by cpu_create() since 2.12 but comments
weren't updated. So update stale comments to point that page
sizes arei actually initialized by tcg_exec_init(). Also move
another qemu_host_page_size related comment before tcg_exec_init()
where it belongs.

Signed-off-by: Igor Mammedov <imammedo@redhat.com>
Reviewed-by: Laurent Vivier <laurent@vivier.eu>
Message-Id: <1526557877-293151-1-git-send-email-imammedo@redhat.com>
Signed-off-by: Laurent Vivier <laurent@vivier.eu>

linux-user: update netlink emulation

Update enums with entries from linux 4.17

Translate entries that generate logs with iproute2 4.9.0 and
host kernel 4.15:

  # ip address show
  Unknown host QEMU_IFLA type: 43
  Unknown host QEMU_IFLA type: 43
  Unknown host QEMU_IFLA type: 43
  Unknown QEMU_IFLA_BR type 41
  Unknown QEMU_IFLA_BR type 42
  Unknown QEMU_IFLA_BR type 43
  Unknown QEMU_IFLA_BR type 44
  Unknown host QEMU_IFLA type: 43
  Unknown QEMU_IFLA_BR type 41
  Unknown QEMU_IFLA_BR type 42
  Unknown QEMU_IFLA_BR type 43
  Unknown QEMU_IFLA_BR type 44
  Unknown host QEMU_IFLA type: 43

Signed-off-by: Laurent Vivier <laurent@vivier.eu>
Message-Id: <20180516221213.11111-1-laurent@vivier.eu>

linux-user: Assert on bad type in thunk_type_align() and thunk_type_size()

In thunk_type_align() and thunk_type_size() we currently return
-1 if the value at the type_ptr isn't one of the TYPE_* values
we understand. However, this should never happen, and if it does
then the calling code will go confusingly wrong because none
of the callsites try to handle an error return. Switch to an
assertion instead, so that if this does somehow happen we'll have
a nice clear backtrace of what happened rather than a weird crash
or misbehaviour.

This also silences various Coverity complaints about not handling
the negative return value (CID 1005735, 1005736, 1005738, 1390582).

Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
Reviewed-by: Laurent Vivier <laurent@vivier.eu>
Message-Id: <20180514174616.19601-1-peter.maydell@linaro.org>
Signed-off-by: Laurent Vivier <laurent@vivier.eu>

Merge remote-tracking branch 'remotes/kraxel/tags/vga-20180524-pull-request' into staging

vga: catch depth 0
hw/display: add new bochs-display device
some cleanups.

# gpg: Signature made Thu 24 May 2018 16:45:46 BST
# gpg:                using RSA key 4CB6D8EED3E87138
# gpg: Good signature from "Gerd Hoffmann (work) <kraxel@redhat.com>"
# gpg:                 aka "Gerd Hoffmann <gerd@kraxel.org>"
# gpg:                 aka "Gerd Hoffmann (private) <kraxel@gmail.com>"
# Primary key fingerprint: A032 8CFF B93A 17A7 9901  FE7D 4CB6 D8EE D3E8 7138

* remotes/kraxel/tags/vga-20180524-pull-request:
  MAINTAINERS: add vga entries
  bochs-display: add pcie support
  bochs-display: add dirty tracking support
  hw/display: add new bochs-display device
  vga-pci: use PCI_VGA_MMIO_SIZE
  vga: move bochs vbe defines to header file
  vga: catch depth 0

Signed-off-by: Peter Maydell <peter.maydell@linaro.org>

docs: tpm: add VM save/restore example and troubleshooting guide

Extend the docs related to TPM with specs related to VM save and
restore and a troubleshooting guide for TPM migration.

Signed-off-by: Stefan Berger <stefanb@linux.vnet.ibm.com>
Reviewed-by: Dr. David Alan Gilbert <dgilbert@redhat.com>

tpm: extend TPM TIS with state migration support

Extend the TPM TIS interface with state migration support.

We need to synchronize with the backend thread to make sure that a command
being processed by the external TPM emulator has completed and its
response been received.

Signed-off-by: Stefan Berger <stefanb@linux.vnet.ibm.com>
Reviewed-by: Dr. David Alan Gilbert <dgilbert@redhat.com>
Reviewed-by: Marc-André Lureau <marcandre.lureau@redhat.com>

tpm: extend TPM emulator with state migration support

Extend the TPM emulator backend device with state migration support.

The external TPM emulator 'swtpm' provides a protocol over
its control channel to retrieve its state blobs. We implement
functions for getting and setting the different state blobs.
In case the setting of the state blobs fails, we return a
negative errno code to fail the start of the VM.

Since we have an external TPM emulator, we need to make sure
that we do not migrate the state for as long as it is busy
processing a request. We need to wait for notification that
the request has completed processing.

Signed-off-by: Stefan Berger <stefanb@linux.vnet.ibm.com>
Reviewed-by: Dr. David Alan Gilbert <dgilbert@redhat.com>
Reviewed-by: Marc-André Lureau <marcandre.lureau@redhat.com>

Merge remote-tracking branch 'remotes/mst/tags/for_upstream' into staging

pc, pci, virtio, vhost: fixes, features

Beginning of merging vDPA, new PCI ID, a new virtio balloon stat, intel
iommu rework fixing a couple of security problems (no CVEs yet), fixes
all over the place.

Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
# gpg: Signature made Wed 23 May 2018 15:41:32 BST
# gpg:                using RSA key 281F0DB8D28D5469
# gpg: Good signature from "Michael S. Tsirkin <mst@kernel.org>"
# gpg:                 aka "Michael S. Tsirkin <mst@redhat.com>"
# Primary key fingerprint: 0270 606B 6F3C DF3D 0B17  0970 C350 3912 AFBE 8E67
#      Subkey fingerprint: 5D09 FD08 71C8 F85B 94CA  8A0D 281F 0DB8 D28D 5469

* remotes/mst/tags/for_upstream: (28 commits)
  intel-iommu: rework the page walk logic
  util: implement simple iova tree
  intel-iommu: trace domain id during page walk
  intel-iommu: pass in address space when page walk
  intel-iommu: introduce vtd_page_walk_info
  intel-iommu: only do page walk for MAP notifiers
  intel-iommu: add iommu lock
  intel-iommu: remove IntelIOMMUNotifierNode
  intel-iommu: send PSI always even if across PDEs
  nvdimm: fix typo in label-size definition
  contrib/vhost-user-blk: enable protocol feature for vhost-user-blk
  hw/virtio: Fix brace Werror with clang 6.0.0
  libvhost-user: Send messages with no data
  vhost-user+postcopy: Use qemu_set_nonblock
  virtio: support setting memory region based host notifier
  vhost-user: support receiving file descriptors in slave_read
  vhost-user: add Net prefix to internal state structure
  linux-headers: add kvm header for mips
  linux-headers: add unistd.h on all arches
  update-linux-headers.sh: unistd.h, kvm consistency
  ...

Signed-off-by: Peter Maydell <peter.maydell@linaro.org>

Merge remote-tracking branch 'remotes/kevin/tags/for-upstream' into staging

Block layer patches:

- Generic background jobs
- qemu-iotests fixes for NFS and the 'migration' group
- sheepdog: Minor code simplification

# gpg: Signature made Wed 23 May 2018 13:33:49 BST
# gpg:                using RSA key 7F09B272C88F2FD6
# gpg: Good signature from "Kevin Wolf <kwolf@redhat.com>"
# Primary key fingerprint: DC3D EB15 9A9A F95D 3D74  56FE 7F09 B272 C88F 2FD6

* remotes/kevin/tags/for-upstream: (46 commits)
  qemu-iotests: Test job-* with block jobs
  iotests: Move qmp_to_opts() to VM
  blockjob: Remove BlockJob.driver
  job: Add query-jobs QMP command
  job: Add lifecycle QMP commands
  job: Add JOB_STATUS_CHANGE QMP event
  job: Introduce qapi/job.json
  job: Move progress fields to Job
  job: Add job_transition_to_ready()
  job: Add job_is_ready()
  job: Add job_dismiss()
  job: Add job_yield()
  block: Cancel job in bdrv_close_all() callers
  job: Move completion and cancellation to Job
  job: Move transactions to Job
  job: Switch transactions to JobTxn
  job: Move job_finish_sync() to Job
  job: Move .complete callback to Job
  job: Add job_drain()
  job: Convert block_job_cancel_async() to Job
  ...

Signed-off-by: Peter Maydell <peter.maydell@linaro.org>

Merge remote-tracking branch 'remotes/sstabellini-http/tags/xen-20180522-tag' into staging

Xen 2018/05/22

# gpg: Signature made Tue 22 May 2018 19:44:06 BST
# gpg:                using RSA key 894F8F4870E1AE90
# gpg: Good signature from "Stefano Stabellini <stefano.stabellini@eu.citrix.com>"
# gpg:                 aka "Stefano Stabellini <sstabellini@kernel.org>"
# Primary key fingerprint: D04E 33AB A51F 67BA 07D3  0AEA 894F 8F48 70E1 AE90

* remotes/sstabellini-http/tags/xen-20180522-tag:
  xen_disk: be consistent with use of xendev and blkdev->xendev
  xen_disk: use a single entry iovec
  xen_backend: make the xen_feature_grant_copy flag private
  xen_disk: remove use of grant map/unmap
  xen_backend: add an emulation of grant copy
  xen: remove other open-coded use of libxengnttab
  xen_disk: remove open-coded use of libxengnttab
  xen_backend: add grant table helpers
  xen: add a meaningful declaration of grant_copy_segment into xen_common.h
  checkpatch: generalize xen handle matching in the list of types
  xen-hvm: create separate function for ioreq server initialization
  xen_pt: Present the size of 64 bit BARs correctly
  configure: Add explanation for --enable-xen-pci-passthrough
  xen/pt: use address_space_memory object for memory region hooks
  xen-pvdevice: Introduce a simplistic xen-pvdevice save state

Signed-off-by: Peter Maydell <peter.maydell@linaro.org>

Merge remote-tracking branch 'remotes/mwalle/tags/lm32-queue/20180521' into staging

target/lm32: BQL patch

# gpg: Signature made Tue 22 May 2018 19:25:30 BST
# gpg:                using RSA key B458ABB0D8D378E3
# gpg: Good signature from "Michael Walle <michael@walle.cc>"
# gpg: WARNING: This key is not certified with a trusted signature!
# gpg:          There is no indication that the signature belongs to the owner.
# Primary key fingerprint: 2190 3E48 4537 A7C2 90CE  3EB2 B458 ABB0 D8D3 78E3

* remotes/mwalle/tags/lm32-queue/20180521:
  lm32: take BQL before writing IP/IM register

Signed-off-by: Peter Maydell <peter.maydell@linaro.org>

MAINTAINERS: add vga entries

Add entries for standard vga, virtio-gpu and cirrus.

Signed-off-by: Gerd Hoffmann <kraxel@redhat.com>
Reviewed-by: Marc-André Lureau <marcandre.lureau@redhat.com>
Reviewed-by: Philippe Mathieu-Daudé <f4bug@amsat.org>
Message-id: 20180522165058.15404-7-kraxel@redhat.com

bochs-display: add pcie support

Signed-off-by: Gerd Hoffmann <kraxel@redhat.com>
Message-id: 20180522165058.15404-6-kraxel@redhat.com

bochs-display: add dirty tracking support

Signed-off-by: Gerd Hoffmann <kraxel@redhat.com>
Reviewed-by: Marc-André Lureau <marcandre.lureau@redhat.com>
Message-id: 20180522165058.15404-5-kraxel@redhat.com

hw/display: add new bochs-display device

After writing up the virtual mdev device emulating a display supporting
the bochs vbe dispi interface (mbochs.ko) and seeing how simple it
actually is I've figured that would be useful for qemu too.

So, here it is, -device bochs-display.  It is basically -device VGA
without legacy vga emulation.  PCI bar 0 is the framebuffer, PCI bar 2
is mmio with the registers.  The vga registers are simply not there
though, neither in the legacy ioport location nor in the mmio bar.
Consequently it is PCI class DISPLAY_OTHER not DISPLAY_VGA.

So there is no text mode emulation, no weird video modes (planar,
256color palette), no memory window at 0xa0000.  Just a linear
framebuffer in the pci memory bar.  And the amount of code to emulate
this (and therefore the attack surface) is an order of magnitude smaller
when compared to vga emulation.

Compatibility wise it works with OVMF (latest git master).
The bochs-drm.ko linux kernel module can handle it just fine too.
So UEFI guests should not see any functional difference to VGA.

Signed-off-by: Gerd Hoffmann <kraxel@redhat.com>
Tested-by: Marc-André Lureau <marcandre.lureau@redhat.com>
Reviewed-by: Marc-André Lureau <marcandre.lureau@redhat.com>
Message-id: 20180522165058.15404-4-kraxel@redhat.com

vga-pci: use PCI_VGA_MMIO_SIZE

Signed-off-by: Gerd Hoffmann <kraxel@redhat.com>
Reviewed-by: Marc-André Lureau <marcandre.lureau@redhat.com>
Reviewed-by: Philippe Mathieu-Daudé <f4bug@amsat.org>
Message-id: 20180522165058.15404-3-kraxel@redhat.com

vga: move bochs vbe defines to header file

Create a new header file, move the bochs vbe dispi interface
defines to it, so they can be used outside vga code.

Signed-off-by: Gerd Hoffmann <kraxel@redhat.com>
Reviewed-by: Philippe Mathieu-Daudé <f4bug@amsat.org>
Reviewed-by: Marc-André Lureau <marcandre.lureau@redhat.com>
Message-id: 20180522165058.15404-2-kraxel@redhat.com

vga: catch depth 0

depth == 0 is used to indicate 256 color modes.  Our region calculation
goes wrong in that case.  So detect that and just take the safe code
path we already have for the wraparound case.

While being at it also catch depth == 15 (where our region size
calculation goes wrong too).  And make the comment more verbose,
explaining what is going on here.

Without this windows guest install might trigger an assert due to trying
to check dirty bitmap outside the snapshot region.

Fixes: https://bugzilla.redhat.com/show_bug.cgi?id=1575541
Signed-off-by: Gerd Hoffmann <kraxel@redhat.com>
Message-id: 20180514103117.21059-1-kraxel@redhat.com

intel-iommu: rework the page walk logic

This patch fixes a potential small window that the DMA page table might
be incomplete or invalid when the guest sends domain/context
invalidations to a device.  This can cause random DMA errors for
assigned devices.

This is a major change to the VT-d shadow page walking logic. It
includes but is not limited to:

- For each VTDAddressSpace, now we maintain what IOVA ranges we have
  mapped and what we have not.  With that information, now we only send
  MAP or UNMAP when necessary.  Say, we don't send MAP notifies if we
  know we have already mapped the range, meanwhile we don't send UNMAP
  notifies if we know we never mapped the range at all.

- Introduce vtd_sync_shadow_page_table[_range] APIs so that we can call
  in any places to resync the shadow page table for a device.

- When we receive domain/context invalidation, we should not really run
  the replay logic, instead we use the new sync shadow page table API to
  resync the whole shadow page table without unmapping the whole
  region.  After this change, we'll only do the page walk once for each
  domain invalidations (before this, it can be multiple, depending on
  number of notifiers per address space).

While at it, the page walking logic is also refactored to be simpler.

CC: QEMU Stable <qemu-stable@nongnu.org>
Reported-by: Jintack Lim <jintack@cs.columbia.edu>
Tested-by: Jintack Lim <jintack@cs.columbia.edu>
Signed-off-by: Peter Xu <peterx@redhat.com>
Reviewed-by: Michael S. Tsirkin <mst@redhat.com>
Signed-off-by: Michael S. Tsirkin <mst@redhat.com>

util: implement simple iova tree

Introduce a simplest iova tree implementation based on GTree.

CC: QEMU Stable <qemu-stable@nongnu.org>
Signed-off-by: Peter Xu <peterx@redhat.com>
Reviewed-by: Michael S. Tsirkin <mst@redhat.com>
Signed-off-by: Michael S. Tsirkin <mst@redhat.com>

intel-iommu: trace domain id during page walk

This patch only modifies the trace points.

Previously we were tracing page walk levels.  They are redundant since
we have page mask (size) already.  Now we trace something much more
useful which is the domain ID of the page walking.  That can be very
useful when we trace more than one devices on the same system, so that
we can know which map is for which domain.

CC: QEMU Stable <qemu-stable@nongnu.org>
Signed-off-by: Peter Xu <peterx@redhat.com>
Reviewed-by: Michael S. Tsirkin <mst@redhat.com>
Signed-off-by: Michael S. Tsirkin <mst@redhat.com>

intel-iommu: pass in address space when page walk

We pass in the VTDAddressSpace too.  It'll be used in the follow up
patches.

CC: QEMU Stable <qemu-stable@nongnu.org>
Signed-off-by: Peter Xu <peterx@redhat.com>
Reviewed-by: Michael S. Tsirkin <mst@redhat.com>
Signed-off-by: Michael S. Tsirkin <mst@redhat.com>

intel-iommu: introduce vtd_page_walk_info

During the recursive page walking of IOVA page tables, some stack
variables are constant variables and never changed during the whole page
walking procedure.  Isolate them into a struct so that we don't need to
pass those contants down the stack every time and multiple times.

CC: QEMU Stable <qemu-stable@nongnu.org>
Signed-off-by: Peter Xu <peterx@redhat.com>
Reviewed-by: Michael S. Tsirkin <mst@redhat.com>
Signed-off-by: Michael S. Tsirkin <mst@redhat.com>

intel-iommu: only do page walk for MAP notifiers

For UNMAP-only IOMMU notifiers, we don't need to walk the page tables.
Fasten that procedure by skipping the page table walk.  That should
boost performance for UNMAP-only notifiers like vhost.

CC: QEMU Stable <qemu-stable@nongnu.org>
Signed-off-by: Peter Xu <peterx@redhat.com>
Reviewed-by: Michael S. Tsirkin <mst@redhat.com>
Signed-off-by: Michael S. Tsirkin <mst@redhat.com>

intel-iommu: add iommu lock

SECURITY IMPLICATION: this patch fixes a potential race when multiple
threads access the IOMMU IOTLB cache.

Add a per-iommu big lock to protect IOMMU status.  Currently the only
thing to be protected is the IOTLB/context cache, since that can be
accessed even without BQL, e.g., in IO dataplane.

Note that we don't need to protect device page tables since that's fully
controlled by the guest kernel.  However there is still possibility that
malicious drivers will program the device to not obey the rule.  In that
case QEMU can't really do anything useful, instead the guest itself will
be responsible for all uncertainties.

CC: QEMU Stable <qemu-stable@nongnu.org>
Reported-by: Fam Zheng <famz@redhat.com>
Signed-off-by: Peter Xu <peterx@redhat.com>
Reviewed-by: Michael S. Tsirkin <mst@redhat.com>
Signed-off-by: Michael S. Tsirkin <mst@redhat.com>

intel-iommu: remove IntelIOMMUNotifierNode

That is not really necessary.  Removing that node struct and put the
list entry directly into VTDAddressSpace.  It simplfies the code a lot.
Since at it, rename the old notifiers_list into vtd_as_with_notifiers.

CC: QEMU Stable <qemu-stable@nongnu.org>
Signed-off-by: Peter Xu <peterx@redhat.com>
Reviewed-by: Michael S. Tsirkin <mst@redhat.com>
Signed-off-by: Michael S. Tsirkin <mst@redhat.com>

intel-iommu: send PSI always even if across PDEs

SECURITY IMPLICATION: without this patch, any guest with both assigned
device and a vIOMMU might encounter stale IO page mappings even if guest
has already unmapped the page, which may lead to guest memory
corruption.  The stale mappings will only be limited to the guest's own
memory range, so it should not affect the host memory or other guests on
the host.

During IOVA page table walking, there is a special case when the PSI
covers one whole PDE (Page Directory Entry, which contains 512 Page
Table Entries) or more.  In the past, we skip that entry and we don't
notify the IOMMU notifiers.  This is not correct.  We should send UNMAP
notification to registered UNMAP notifiers in this case.

For UNMAP only notifiers, this might cause IOTLBs cached in the devices
even if they were already invalid.  For MAP/UNMAP notifiers like
vfio-pci, this will cause stale page mappings.

This special case doesn't trigger often, but it is very easy to be
triggered by nested device assignments, since in that case we'll
possibly map the whole L2 guest RAM region into the device's IOVA
address space (several GBs at least), which is far bigger than normal
kernel driver usages of the device (tens of MBs normally).

Without this patch applied to L1 QEMU, nested device assignment to L2
guests will dump some errors like:

qemu-system-x86_64: VFIO_MAP_DMA: -17
qemu-system-x86_64: vfio_dma_map(0x557305420c30, 0xad000, 0x1000,
                    0x7f89a920d000) = -17 (File exists)

CC: QEMU Stable <qemu-stable@nongnu.org>
Acked-by: Jason Wang <jasowang@redhat.com>
[peterx: rewrite the commit message]
Signed-off-by: Peter Xu <peterx@redhat.com>
Reviewed-by: Michael S. Tsirkin <mst@redhat.com>
Signed-off-by: Michael S. Tsirkin <mst@redhat.com>

nvdimm: fix typo in label-size definition

Signed-off-by: Ross Zwisler <ross.zwisler@linux.intel.com>
Fixes: commit da6789c27c2e ("nvdimm: add a macro for property "label-size"")
Reviewed-by: Stefan Hajnoczi <stefanha@redhat.com>
Reviewed-by: Igor Mammedov <imammedo@redhat.com>
Cc: Haozhong Zhang <haozhong.zhang@intel.com>
Cc: Michael S. Tsirkin <mst@redhat.com>
Cc: Stefan Hajnoczi <stefanha@redhat.com>
Reviewed-by: Michael S. Tsirkin <mst@redhat.com>
Signed-off-by: Michael S. Tsirkin <mst@redhat.com>

contrib/vhost-user-blk: enable protocol feature for vhost-user-blk

This patch reports the protocol feature that is only advertised by
QEMU if the device implements the config ops.

Signed-off-by: Changpeng Liu <changpeng.liu@intel.com>
Reviewed-by: Stefan Hajnoczi <stefanha@redhat.com>
Reviewed-by: Michael S. Tsirkin <mst@redhat.com>
Signed-off-by: Michael S. Tsirkin <mst@redhat.com>

hw/virtio: Fix brace Werror with clang 6.0.0

The warning is

hw/virtio/vhost-user.c:1319:26: error: suggest braces
      around initialization of subobject [-Werror,-Wmissing-braces]
    VhostUserMsg msg = { 0 };
                         ^
                         {}

While the original code is correct, and technically exactly correct
as per ISO C89, both GCC and Clang support plain empty set of braces
as an extension.

Cc: Michael S. Tsirkin <mst@redhat.com>
Signed-off-by: Richard Henderson <richard.henderson@linaro.org>
Reviewed-by: Philippe Mathieu-Daudé <f4bug@amsat.org>
Reviewed-by: Eric Blake <eblake@redhat.com>
Reviewed-by: Michael S. Tsirkin <mst@redhat.com>
Signed-off-by: Michael S. Tsirkin <mst@redhat.com>

libvhost-user: Send messages with no data

The response to a VHOST_USER_POSTCOPY_ADVISE contains a fd but doesn't
actually contain any data.   FIx vu_message_write so that it doesn't
do a 0-byte write() call, since this was ending up with rc=0
that was confusing the error handling code.

Signed-off-by: Dr. David Alan Gilbert <dgilbert@redhat.com>
Reviewed-by: Marc-André Lureau <marcandre.lureau@redhat.com>
Reviewed-by: Michael S. Tsirkin <mst@redhat.com>
Signed-off-by: Michael S. Tsirkin <mst@redhat.com>

vhost-user+postcopy: Use qemu_set_nonblock

Use qemu_set_nonblock rather than a simple fcntl; cleaner
and I have no reason to change other flags.

Reported-by: Peter Maydell <peter.maydell@linaro.org>
Signed-off-by: Dr. David Alan Gilbert <dgilbert@redhat.com>
Reviewed-by: Eric Blake <eblake@redhat.com>
Reviewed-by: Stefan Hajnoczi <stefanha@redhat.com>
Reviewed-by: Michael S. Tsirkin <mst@redhat.com>
Signed-off-by: Michael S. Tsirkin <mst@redhat.com>

virtio: support setting memory region based host notifier

This patch introduces the support for setting memory region
based host notifiers for virtio device. This is helpful when
using a hardware accelerator for a virtio device, because
hardware heavily depends on the notification, this will allow
the guest driver in the VM to notify the hardware directly.

Signed-off-by: Tiwei Bie <tiwei.bie@intel.com>
Reviewed-by: Michael S. Tsirkin <mst@redhat.com>
Signed-off-by: Michael S. Tsirkin <mst@redhat.com>

vhost-user: support receiving file descriptors in slave_read

Signed-off-by: Tiwei Bie <tiwei.bie@intel.com>
Reviewed-by: Michael S. Tsirkin <mst@redhat.com>
Signed-off-by: Michael S. Tsirkin <mst@redhat.com>