tests: qtest: add qtest_has_accel() to check if tested binary supports accelerator

Currently it is not possible to create tests that have KVM as a hard
requirement on a host that doesn't support KVM for tested target
binary (modulo going through the trouble of compiling out
the offending test case).

Following scenario makes test fail when it's run on non x86 host:
  qemu-system-x86_64 -enable-kvm -M q35,kernel-irqchip=on -smp 1,maxcpus=288

This patch introduces qtest_has_accel() to let users check if accel is
available in advance and avoid executing non run-able test-cases.

It implements detection of TCG and KVM only, the rest could be
added later on, when we actually start testing them in qtest.

Signed-off-by: Igor Mammedov <imammedo@redhat.com>
Message-Id: <20210902113551.461632-3-imammedo@redhat.com>
Reviewed-by: Michael S. Tsirkin <mst@redhat.com>
Signed-off-by: Michael S. Tsirkin <mst@redhat.com>

tests: acpi: dump table with failed checksum

Signed-off-by: Igor Mammedov <imammedo@redhat.com>
Message-Id: <20210902113551.461632-2-imammedo@redhat.com>
Reviewed-by: Michael S. Tsirkin <mst@redhat.com>
Signed-off-by: Michael S. Tsirkin <mst@redhat.com>

Merge remote-tracking branch 'remotes/juanquintela/tags/migration.next-pull-request' into staging

Migration Pull request (3rd try)

Hi

This should fix all the freebsd problems.

Please apply,

# gpg: Signature made Tue 19 Oct 2021 02:28:51 AM PDT
# gpg:                using RSA key 1899FF8EDEBF58CCEE034B82F487EF185872D723
# gpg: Good signature from "Juan Quintela <quintela@redhat.com>" [full]
# gpg:                 aka "Juan Quintela <quintela@trasno.org>" [full]

* remotes/juanquintela/tags/migration.next-pull-request:
  migration/rdma: advise prefetch write for ODP region
  migration/rdma: Try to register On-Demand Paging memory region
  migration: allow enabling mutilfd for specific protocol only
  migration: allow multifd for socket protocol only
  migration/ram: Don't passs RAMState to migration_clear_memory_region_dirty_bitmap_*()
  multifd: Unconditionally unregister yank function
  multifd: Implement yank for multifd send side

Signed-off-by: Richard Henderson <richard.henderson@linaro.org>

migration/rdma: advise prefetch write for ODP region

The responder mr registering with ODP will sent RNR NAK back to
the requester in the face of the page fault.
---------
ibv_poll_cq wc.status=13 RNR retry counter exceeded!
ibv_poll_cq wrid=WRITE RDMA!
---------
ibv_advise_mr(3) helps to make pages present before the actual IO is
conducted so that the responder does page fault as little as possible.

Signed-off-by: Li Zhijian <lizhijian@cn.fujitsu.com>
Reviewed-by: Marcel Apfelbaum <marcel.apfelbaum@gmail.com>
Reviewed-by: Juan Quintela <quintela@redhat.com>
Signed-off-by: Juan Quintela <quintela@redhat.com>

migration/rdma: Try to register On-Demand Paging memory region

Previously, for the fsdax mem-backend-file, it will register failed with
Operation not supported. In this case, we can try to register it with
On-Demand Paging[1] like what rpma_mr_reg() does on rpma[2].

[1]: https://community.mellanox.com/s/article/understanding-on-demand-paging--odp-x
[2]: http://pmem.io/rpma/manpages/v0.9.0/rpma_mr_reg.3

CC: Marcel Apfelbaum <marcel.apfelbaum@gmail.com>
Signed-off-by: Li Zhijian <lizhijian@cn.fujitsu.com>
Reviewed-by: Marcel Apfelbaum <marcel.apfelbaum@gmail.com>
Reviewed-by: Juan Quintela <quintela@redhat.com>
Signed-off-by: Juan Quintela <quintela@redhat.com>

migration: allow enabling mutilfd for specific protocol only

To: <quintela@redhat.com>, <dgilbert@redhat.com>, <qemu-devel@nongnu.org>
CC: Li Zhijian <lizhijian@cn.fujitsu.com>
Date: Sat, 31 Jul 2021 22:05:52 +0800 (5 weeks, 4 days, 17 hours ago)

And change the default to true so that in '-incoming defer' case, user is able
to change multifd capability.

Signed-off-by: Li Zhijian <lizhijian@cn.fujitsu.com>
Reviewed-by: Juan Quintela <quintela@redhat.com>
Signed-off-by: Juan Quintela <quintela@redhat.com>

migration: allow multifd for socket protocol only

To: <quintela@redhat.com>, <dgilbert@redhat.com>, <qemu-devel@nongnu.org>
CC: Li Zhijian <lizhijian@cn.fujitsu.com>
Date: Sat, 31 Jul 2021 22:05:51 +0800 (5 weeks, 4 days, 17 hours ago)

multifd with unsupported protocol will cause a segment fault.
(gdb) bt
 #0  0x0000563b4a93faf8 in socket_connect (addr=0x0, errp=0x7f7f02675410) at ../util/qemu-sockets.c:1190
 #1 0x0000563b4a797a03 in qio_channel_socket_connect_sync
(ioc=0x563b4d16e8c0, addr=0x0, errp=0x7f7f02675410) at
../io/channel-socket.c:145
 #2  0x0000563b4a797abf in qio_channel_socket_connect_worker (task=0x563b4cd86c30, opaque=0x0) at ../io/channel-socket.c:168
 #3  0x0000563b4a792631 in qio_task_thread_worker (opaque=0x563b4cd86c30) at ../io/task.c:124
 #4  0x0000563b4a91da69 in qemu_thread_start (args=0x563b4c44bb80) at ../util/qemu-thread-posix.c:541
 #5  0x00007f7fe9b5b3f9 in ?? ()
 #6  0x0000000000000000 in ?? ()

It's enough to check migrate_multifd_is_allowed() in multifd cleanup() and
multifd setup() though there are so many other places using migrate_use_multifd().

Signed-off-by: Li Zhijian <lizhijian@cn.fujitsu.com>
Reviewed-by: Juan Quintela <quintela@redhat.com>
Signed-off-by: Juan Quintela <quintela@redhat.com>

migration/ram: Don't passs RAMState to migration_clear_memory_region_dirty_bitmap_*()

The parameter is unused, let's drop it.

Reviewed-by: Peter Xu <peterx@redhat.com>
Reviewed-by: Philippe Mathieu-Daudé <philmd@redhat.com>
Reviewed-by: Juan Quintela <quintela@redhat.com>
Signed-off-by: David Hildenbrand <david@redhat.com>
Signed-off-by: Juan Quintela <quintela@redhat.com>

multifd: Unconditionally unregister yank function

To: qemu-devel <qemu-devel@nongnu.org>
Cc: "Dr. David Alan Gilbert" <dgilbert@redhat.com>, Juan Quintela
 <quintela@redhat.com>, Peter Xu <peterx@redhat.com>, Leonardo Bras Soares
 Passos <lsoaresp@redhat.com>
Date: Wed, 4 Aug 2021 21:26:32 +0200 (5 weeks, 11 hours, 52 minutes ago)

[[PGP Signed Part:No public key for 35AB0B289C5DB258 created at 2021-08-04T21:26:32+0200 using RSA]]
Unconditionally unregister yank function in multifd_load_cleanup().
If it is not unregistered here, it will leak and cause a crash
in yank_unregister_instance(). Now if the ioc is still in use
afterwards, it will only lead to qemu not being able to recover
from a hang related to that ioc.

After checking the code, i am pretty sure that ref is always 1
when arriving here. So all this currently does is remove the
unneeded check.

Signed-off-by: Lukas Straub <lukasstraub2@web.de>
Reviewed-by: Juan Quintela <quintela@redhat.com>
Signed-off-by: Juan Quintela <quintela@redhat.com>

multifd: Implement yank for multifd send side

To: qemu-devel <qemu-devel@nongnu.org>
Cc: "Dr. David Alan Gilbert" <dgilbert@redhat.com>, Juan Quintela
 <quintela@redhat.com>, Peter Xu <peterx@redhat.com>, Leonardo Bras Soares
 Passos <lsoaresp@redhat.com>
Date: Wed, 1 Sep 2021 17:58:57 +0200 (1 week, 15 hours, 17 minutes ago)

[[PGP Signed Part:No public key for 35AB0B289C5DB258 created at 2021-09-01T17:58:57+0200 using RSA]]
When introducing yank functionality in the migration code I forgot
to cover the multifd send side.

Signed-off-by: Lukas Straub <lukasstraub2@web.de>
Tested-by: Leonardo Bras <leobras@redhat.com>
Reviewed-by: Leonardo Bras <leobras@redhat.com>
Reviewed-by: Juan Quintela <quintela@redhat.com>
Signed-off-by: Juan Quintela <quintela@redhat.com>

Merge remote-tracking branch 'remotes/bsdimp/tags/pull-bsd-user-20211018-pull-request' into staging

bsd-user pull request: merge dependencies for next architectures

Merge the dependencies for arm, aarch64, and riscv64 architectures. This joins
together two patch series:

[PATCH v2 00/15] bsd-user: misc cleanup for aarch64 import

Prepare for aarch64 support (the next architecture to be upstreamed). As the
aarch64 emulation is more complete, it relies on a number of different items.
In some cases, I've pulled in the full support from bsd-user fork. In other
cases I've created a simple stub (as is the case for signals, which have
independent changes pending, so I wanted to be as minimal as possible.  Since
all pre-12.2 support was purged from the bsd-user fork, go ahead and remove it
here. FreeBSD 11.x goes ouft of support at the end of the month. Remove what
little multi-version support that's in upstream.

and

[PATCH v3 0/9] bsd-user mmap fixes
This series synchronizes mmap.c with the bsd-user fork. This is a mix of old bug
fixes pulled in from linux-user, as well as some newer fixes to adress bugs
found in check-tcg and recent FreeBSD developments. There are also a couple of
style commits. Updated to migrate debugging to qemu_log.

as well as a couple of minor rebase tweaks. In addition, the next two
architectures I plan on upstreaming (arm and riscv64) also have their prereqs
satisfied with this request.

v2: Remove accidental module regression in patch 7 and try again.

# gpg: Signature made Mon 18 Oct 2021 12:00:28 PM PDT
# gpg:                using RSA key 2035F894B00AA3CF7CCDE1B76C1CD1287DB01100
# gpg: Good signature from "Warner Losh <wlosh@netflix.com>" [unknown]
# gpg:                 aka "Warner Losh <imp@bsdimp.com>" [unknown]
# gpg:                 aka "Warner Losh <imp@freebsd.org>" [unknown]
# gpg:                 aka "Warner Losh <imp@village.org>" [unknown]
# gpg:                 aka "Warner Losh <wlosh@bsdimp.com>" [unknown]
# 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: 2035 F894 B00A A3CF 7CCD  E1B7 6C1C D128 7DB0 1100

* remotes/bsdimp/tags/pull-bsd-user-20211018-pull-request: (23 commits)
  bsd-user/signal: Create a dummy signal queueing function
  bsd-user: Rename sigqueue to qemu_sigqueue
  bsd-user/sysarch: Move to using do_freebsd_arch_sysarch interface
  bsd-user: Add stop_all_tasks
  bsd-user: Remove used from TaskState
  bsd-user/target_os_elf: If ELF_HWCAP2 is defined, publish it
  bsd-user/target_os_elf.h: Remove fallback ELF_HWCAP and reorder
  bsd-user: move TARGET_MC_GET_CLEAR_RET to target_os_signal.h
  bsd-user/errno_defs.h: Add internal error numbers
  bsd-user: export get_errno and is_error from syscall.c
  bsd-user: TARGET_RESET define is unused, remove it
  bsd-user/strace.list: Remove support for FreeBSD versions older than 12.0
  bsd-user/target_os-user.h: Remove support for FreeBSD older than 12.0
  meson: *-user: only descend into *-user when configured
  bsd-user/mmap.c: assert that target_mprotect cannot fail
  bsd-user/mmap.c: Implement MAP_EXCL, required by jemalloc in head
  bsd-user/mmap.c: Don't mmap fd == -1 independently from MAP_ANON flag
  bsd-user/mmap.c: Convert to qemu_log logging for mmap debugging
  bsd-user/mmap.c: mmap prefer MAP_ANON for BSD
  bsd-user/mmap.c: mmap return ENOMEM on overflow
  ...

Signed-off-by: Richard Henderson <richard.henderson@linaro.org>

bsd-user/signal: Create a dummy signal queueing function

Create dummy signal queueing function so we can start to integrate other
architectures (at the cost of signals remaining broken) to tame the
dependency graph a bit and to bring in signals in a more controlled
fashion.  Log unimplemented events to it in the mean time.

Signed-off-by: Warner Losh <imp@bsdimp.com>
Reviewed-by: Richard Henderson <richard.henderson@linaro.org>
Reviewed-by: Kyle Evans <kevans@FreeBSD.org>

bsd-user: Rename sigqueue to qemu_sigqueue

To avoid a name clash with FreeBSD's sigqueue data structure in
signalvar.h, rename sigqueue to qemu_sigqueue. This structure
is currently defined, but unused.

Signed-off-by: Warner Losh <imp@bsdimp.com>
Reviewed-by: Richard Henderson <richard.henderson@linaro.org>
Reviewed-by: Philippe Mathieu-Daudé <f4bug@amsat.org>
Reviewed-by: Kyle Evans <kevans@FreeBSD.org>

bsd-user/sysarch: Move to using do_freebsd_arch_sysarch interface

do_freebsd_arch_sysarch() exists in $ARCH/target_arch_sysarch.h for x86.
Call it from do_freebsd_sysarch() and remove the mostly duplicate
version in syscall.c. Future changes will move it to os-sys.c and
support other architectures.

Signed-off-by: Warner Losh <imp@bsdimp.com>
Reviewed-by: Richard Henderson <richard.henderson@linaro.org>
Reviewed-by: Kyle Evans <kevans@FreeBSD.org>

bsd-user: Add stop_all_tasks

Similar to the same function in linux-user: this stops all the current tasks.

Signed-off-by: Stacey Son <sson@FreeBSD.org>
Signed-off-by: Warner Losh <imp@bsdimp.com>
Reviewed-by: Richard Henderson <richard.henderson@linaro.org>
Reviewed-by: Kyle Evans <kevans@FreeBSD.org>

bsd-user: Remove used from TaskState

The 'used' field in TaskState is write only. Remove it from TaskState.

Signed-off-by: Warner Losh <imp@bsdimp.com>
Reviewed-by: Richard Henderson <richard.henderson@linaro.org>
Reviewed-by: Philippe Mathieu-Daudé <f4bug@amsat.org>
Reviewed-by: Kyle Evans <kevans@FreeBSD.org>

bsd-user/target_os_elf: If ELF_HWCAP2 is defined, publish it

Some architectures publish AT_HWCAP2 as well as AT_HWCAP. Those
architectures will define ELF_HWCAP2 in their target_arch_elf.h files
for the value for this process. If it is defined, then publish it.

Signed-off-by: Warner Losh <imp@bsdimp.com>
Reviewed-by: Richard Henderson <richard.henderson@linaro.org>
Reviewed-by: Philippe Mathieu-Daudé <f4bug@amsat.org>
Reviewed-by: Kyle Evans <kevans@FreeBSD.org>

bsd-user/target_os_elf.h: Remove fallback ELF_HWCAP and reorder

All architectures have a ELF_HWCAP, so remove the fallback ifdef.
Place ELF_HWCAP in the same order as on native FreeBSD.

Signed-off-by: Warner Losh <imp@bsdimp.com>
Reviewed-by: Richard Henderson <richard.henderson@linaro.org>
Reviewed-by: Philippe Mathieu-Daudé <f4bug@amsat.org>
Reviewed-by: Kyle Evans <kevans@FreeBSD.org>

bsd-user: move TARGET_MC_GET_CLEAR_RET to target_os_signal.h

Move TARGET_MC_GET_CLEAR_RET to freebsd/target_os_signal.h since it's
architecture agnostic on FreeBSD.

Signed-off-by: Warner Losh <imp@bsdimp.com>
Reviewed-by: Richard Henderson <richard.henderson@linaro.org>
Reviewed-by: Philippe Mathieu-Daudé <f4bug@amsat.org>
Reviewed-by: Kyle Evans <kevans@FreeBSD.org>

bsd-user/errno_defs.h: Add internal error numbers

To emulate signals and interrupted system calls, we need to have the
same mechanisms we have in the kernel, including these errno values.

Signed-off-by: Stacey Son <sson@FreeBSD.org>
Signed-off-by: Warner Losh <imp@bsdimp.com>
Reviewed-by: Richard Henderson <richard.henderson@linaro.org>
Reviewed-by: Kyle Evans <kevans@FreeBSD.org>

bsd-user: export get_errno and is_error from syscall.c

Make get_errno and is_error global so files other than syscall.c can use
them.

Signed-off-by: Warner Losh <imp@bsdimp.com>
Reviewed-by: Richard Henderson <richard.henderson@linaro.org>
Reviewed-by: Philippe Mathieu-Daudé <f4bug@amsat.org>
Reviewed-by: Kyle Evans <kevans@FreeBSD.org>

bsd-user: TARGET_RESET define is unused, remove it

Signed-off-by: Warner Losh <imp@bsdimp.com>
Reviewed-by: Richard Henderson <richard.henderson@linaro.org>
Reviewed-by: Philippe Mathieu-Daudé <f4bug@amsat.org>
Reviewed-by: Kyle Evans <kevans@FreeBSD.org>

bsd-user/strace.list: Remove support for FreeBSD versions older than 12.0

Signed-off-by: Warner Losh <imp@bsdimp.com>
Reviewed-by: Richard Henderson <richard.henderson@linaro.org>
Reviewed-by: Kyle Evans <kevans@FreeBSD.org>

bsd-user/target_os-user.h: Remove support for FreeBSD older than 12.0

Signed-off-by: Warner Losh <imp@bsdimp.com>
Reviewed-by: Richard Henderson <richard.henderson@linaro.org>
Reviewed-by: Kyle Evans <kevans@FreeBSD.org>

meson: *-user: only descend into *-user when configured

To increase flexibility, only descend into *-user when that is
configured. This allows *-user to selectively include directories based
on the host OS which may not exist on all hosts. Adopt Paolo's
suggestion of checking the configuration in the directories that know
about the configuration.

Message-Id: <20210926220103.1721355-2-f4bug@amsat.org>
Message-Id: <20210926220103.1721355-3-f4bug@amsat.org>
Signed-off-by: Philippe Mathieu-Daudé <f4bug@amsat.org>
Signed-off-by: Warner Losh <wlosh@bsdimp.com>
Acked-by: Paolo Bonzini <pbonzinni@redhat.com>
Reviewed-by: Kyle Evans <kevans@FreeBSD.org>

bsd-user/mmap.c: assert that target_mprotect cannot fail

Similar to the equivalent linux-user change 86abac06c14. All error
conditions that target_mprotect checks are also checked by target_mmap.
EACCESS cannot happen because we are just removing PROT_WRITE.  ENOMEM
should not happen because we are modifying a whole VMA (and we have
bigger problems anyway if it happens).

Fixes a Coverity false positive, where Coverity complains about
target_mprotect's return value being passed to tb_invalidate_phys_range.

Signed-off-by: Mikaël Urankar <mikael.urankar@gmail.com>
Signed-off-by: Warner Losh <imp@bsdimp.com>
Reviewed-by: Richard Henderson <richard.henderson@linaro.org>
Reviewed-by: Philippe Mathieu-Daudé <f4bug@amsat.org>
Reviewed-by: Kyle Evans <kevans@FreeBSD.org>

bsd-user/mmap.c: Implement MAP_EXCL, required by jemalloc in head

jemalloc requires a working MAP_EXCL. Ensure that no page is double
mapped when specified. In addition, use guest_range_valid_untagged to
test for valid ranges of pages rather than an incomplete inlined version
of the test that might be wrong.

Signed-off-by: Kyle Evans <kevans@FreeBSD.org>
Signed-off-by: Warner Losh <imp@bsdimp.com>
Reviewed-by: Richard Henderson <richard.henderson@linaro.org>

bsd-user/mmap.c: Don't mmap fd == -1 independently from MAP_ANON flag

Switch checks for !(flags & MAP_ANONYMOUS) with checks for fd != -1.
MAP_STACK and MAP_GUARD both require fd == -1 and don't require mapping
the fd either. Add analysis from Guy Yur detailing the different cases
for MAP_GUARD and MAP_STACK.

Signed-off-by: Guy Yur <guyyur@gmail.com>
[ partially merged before, finishing the job and documenting origin]
Signed-off-by: Warner Losh <imp@bsdimp.com>
Reviewed-by: Richard Henderson <richard.henderson@linaro.org>
Reviewed-by: Kyle Evans <kevans@FreeBSD.org>

Merge remote-tracking branch 'remotes/philmd/tags/mips-20211018' into staging

MIPS patches queue

Hardware emulation:
- Generate FDT blob for Boston machine (Jiaxun)
- VIA chipset cleanups (Zoltan)

TCG:
- Use tcg_constant() in Compact branch and MSA opcodes
- Restrict nanoMIPS DSP MULT[U] opcode accumulator to Rel6
- Fix DEXTRV_S.H DSP opcode
- Remove unused TCG temporary for some DSP opcodes

# gpg: Signature made Sun 17 Oct 2021 03:50:57 PM PDT
# gpg:                using RSA key FAABE75E12917221DCFD6BB2E3E32C2CDEADC0DE
# gpg: Good signature from "Philippe Mathieu-Daudé (F4BUG) <f4bug@amsat.org>" [full]

* remotes/philmd/tags/mips-20211018:
  via-ide: Avoid using isa_get_irq()
  vt82c686: Add a method to VIA_ISA to raise ISA interrupts
  vt82c686: Move common code to via_isa_realize
  via-ide: Set user_creatable to false
  target/mips: Remove unused TCG temporary in gen_mipsdsp_accinsn()
  target/mips: Fix DEXTRV_S.H DSP opcode
  target/mips: Use tcg_constant_tl() in gen_compute_compact_branch()
  target/mips: Use explicit extract32() calls in gen_msa_i5()
  target/mips: Use tcg_constant_i32() in gen_msa_3rf()
  target/mips: Use tcg_constant_i32() in gen_msa_2r()
  target/mips: Use tcg_constant_i32() in gen_msa_2rf()
  target/mips: Use tcg_constant_i32() in gen_msa_elm_df()
  target/mips: Remove unused register from MSA 2R/2RF instruction format
  hw/mips/boston: Add FDT generator
  hw/mips/boston: Allow loading elf kernel and dtb
  hw/mips/boston: Massage memory map information
  target/mips: Check nanoMIPS DSP MULT[U] accumulator with Release 6

Signed-off-by: Richard Henderson <richard.henderson@linaro.org>

bsd-user/mmap.c: Convert to qemu_log logging for mmap debugging

Convert DEBUG_MMAP to qemu_log CPU_LOG_PAGE.

Signed-off-by: Warner Losh <imp@bsdimp.com>
Reviewed-by: Richard Henderson <richard.henderson@linaro.org>
Reviewed-by: Philippe Mathieu-Daudé <f4bug@amsat.org>
Reviewed-by: Kyle Evans <kevans@FreeBSD.org>

bsd-user/mmap.c: mmap prefer MAP_ANON for BSD

MAP_ANON and MAP_ANONYMOUS are identical. Prefer MAP_ANON for BSD since
the file is now a confusing mix of the two.

Signed-off-by: Warner Losh <imp@bsdimp.com>
Reviewed-by: Philippe Mathieu-Daudé <f4bug@amsat.org>
Reviewed-by: Richard Henderson <richard.henderson@linaro.org>
Reviewed-by: Kyle Evans <kevans@FreeBSD.org>

bsd-user/mmap.c: mmap return ENOMEM on overflow

mmap should return ENOMEM on len overflow rather than EINVAL. Return
EINVAL when len == 0 and ENOMEM when the rounded to a page length is 0.
Found by make check-tcg.

Signed-off-by: Warner Losh <imp@bsdimp.com>
Reviewed-by: Richard Henderson <richard.henderson@linaro.org>
Reviewed-by: Philippe Mathieu-Daudé <f4bug@amsat.org>
Reviewed-by: Kyle Evans <kevans@FreeBSD.org>

bsd-user/mmap.c: MAP_ symbols are defined, so no need for ifdefs

All these MAP_ symbols are always defined on supported FreeBSD versions
(12.2 and newer), so remove the #ifdefs since they aren't needed.

Signed-off-by: Warner Losh <imp@bsdimp.com>
Reviewed-by: Philippe Mathieu-Daudé <f4bug@amsat.org>
Acked-by: Richard Henderson <richard.henderson@linaro.org>
Reviewed-by: Kyle Evans <kevans@FreeBSD.org>

bsd-user/mmap.c: check pread's return value to fix warnings with _FORTIFY_SOURCE

Simmilar to the equivalent linux-user: commit fb7e378cf9c, which added
checking to pread's return value. Update to current qemu standards with
{} around the if statement.

Signed-off-by: Mikaël Urankar <mikael.urankar@gmail.com>
Signed-off-by: Warner Losh <imp@bsdimp.com>
Reviewed-by: Richard Henderson <richard.henderson@linaro.org>
Reviewed-by: Philippe Mathieu-Daudé <f4bug@amsat.org>
Reviewed-by: Kyle Evans <kevans@FreeBSD.org>

bsd-user/mmap.c: Always zero MAP_ANONYMOUS memory in mmap_frag()

Similar to the equivalent linux-user commit e6deac9cf99

When mapping MAP_ANONYMOUS memory fragments, still need notice about to
set it zero, or it will cause issues.

Signed-off-by: Mikaël Urankar <mikael.urankar@gmail.com>
Signed-off-by: Warner Losh <imp@bsdimp.com>
Reviewed-by: Richard Henderson <richard.henderson@linaro.org>
Reviewed-by: Kyle Evans <kevans@FreeBSD.org>

via-ide: Avoid using isa_get_irq()

Use via_isa_set_irq() which better encapsulates irq handling in the
vt82xx model and avoids using isa_get_irq() that has a comment saying
it should not be used.

Signed-off-by: BALATON Zoltan <balaton@eik.bme.hu>
Reviewed-by: Philippe Mathieu-Daudé <f4bug@amsat.org>
Message-Id: <26cb1848c9fc0360df7a57c2c9ba5e03c4a692b5.1634259980.git.balaton@eik.bme.hu>
Signed-off-by: Philippe Mathieu-Daudé <f4bug@amsat.org>

vt82c686: Add a method to VIA_ISA to raise ISA interrupts

Other functions in the VT82xx chips need to raise ISA interrupts. Keep
a reference to them in the device state and add via_isa_set_irq() to
allow setting their state.

Signed-off-by: BALATON Zoltan <balaton@eik.bme.hu>
Reviewed-by: Philippe Mathieu-Daudé <f4bug@amsat.org>
Reviewed-by: Jiaxun Yang <jiaxun.yang@flygoat.com>
Message-Id: <778c04dc2c8affac060b8edf9e8d7dab3c3e04eb.1634259980.git.balaton@eik.bme.hu>
Signed-off-by: Philippe Mathieu-Daudé <f4bug@amsat.org>

vt82c686: Move common code to via_isa_realize

The vt82c686b_realize and vt8231_realize methods are almost identical,
factor out the common parts to a via_isa_realize function to avoid
code duplication.

Signed-off-by: BALATON Zoltan <balaton@eik.bme.hu>
Reviewed-by: Jiaxun Yang <jiaxun.yang@flygoat.com>
Reviewed-by: Philippe Mathieu-Daudé <f4bug@amsat.org>
Message-Id: <7cb7a16ff4daf8f48d576246255bea1fd355207c.1634259980.git.balaton@eik.bme.hu>
Signed-off-by: Philippe Mathieu-Daudé <f4bug@amsat.org>

via-ide: Set user_creatable to false

This model only works as a function of the via superio chip not as a
standalone PCI device.

Signed-off-by: BALATON Zoltan <balaton@eik.bme.hu>
Reviewed-by: Philippe Mathieu-Daudé <f4bug@amsat.org>
Message-Id: <20211015092159.3E863748F57@zero.eik.bme.hu>
Signed-off-by: Philippe Mathieu-Daudé <f4bug@amsat.org>

target/mips: Remove unused TCG temporary in gen_mipsdsp_accinsn()

Since gen_mipsdsp_accinsn() got added in commit b53371ed5d4
("target-mips: Add ASE DSP accumulator instructions"), the
'v2_t' TCG temporary has never been used. Remove it.

Signed-off-by: Philippe Mathieu-Daudé <f4bug@amsat.org>
Reviewed-by: Richard Henderson <richard.henderson@linaro.org>
Message-Id: <20211014224551.2204949-1-f4bug@amsat.org>

target/mips: Fix DEXTRV_S.H DSP opcode

While for the DEXTR_S.H opcode:

  "The shift argument is provided in the instruction."

For the DEXTRV_S.H opcode we have:

  "The five least-significant bits of register rs provide the
   shift argument, interpreted as a five-bit unsigned integer;
   the remaining bits in rs are ignored."

While 't1' contains the 'rs' register content (the shift value
for DEXTR_S.H), we need to load the value of 'rs' for DEXTRV_S.H.
We can directly use the v1_t TCG register which already contains
this shift value.

Fixes: b53371ed5d4 ("target-mips: Add ASE DSP accumulator instructions")
Signed-off-by: Philippe Mathieu-Daudé <f4bug@amsat.org>
Reviewed-by: Richard Henderson <richard.henderson@linaro.org>
Message-Id: <20211013215652.1764551-1-f4bug@amsat.org>

target/mips: Use tcg_constant_tl() in gen_compute_compact_branch()

The offset is constant and read-only: move it to the constant pool.

Signed-off-by: Philippe Mathieu-Daudé <f4bug@amsat.org>
Reviewed-by: Richard Henderson <richard.henderson@linaro.org>
Message-Id: <20211003175743.3738710-9-f4bug@amsat.org>

target/mips: Use explicit extract32() calls in gen_msa_i5()

We already use sextract32(), use extract32() for completeness
instead of open-coding it.

Signed-off-by: Philippe Mathieu-Daudé <f4bug@amsat.org>
Reviewed-by: Richard Henderson <richard.henderson@linaro.org>
Message-Id: <20211003175743.3738710-7-f4bug@amsat.org>

target/mips: Use tcg_constant_i32() in gen_msa_3rf()

Avoid using a TCG temporary by moving Data Format to the constant pool.

Signed-off-by: Philippe Mathieu-Daudé <f4bug@amsat.org>
Reviewed-by: Richard Henderson <richard.henderson@linaro.org>
Message-Id: <20211003175743.3738710-6-f4bug@amsat.org>

target/mips: Use tcg_constant_i32() in gen_msa_2r()

Avoid using a TCG temporary by moving Data Format to the constant pool.

Signed-off-by: Philippe Mathieu-Daudé <f4bug@amsat.org>
Reviewed-by: Richard Henderson <richard.henderson@linaro.org>
Message-Id: <20211003175743.3738710-5-f4bug@amsat.org>

target/mips: Use tcg_constant_i32() in gen_msa_2rf()

Avoid using a TCG temporary by moving Data Format to the constant pool.

Signed-off-by: Philippe Mathieu-Daudé <f4bug@amsat.org>
Reviewed-by: Richard Henderson <richard.henderson@linaro.org>
Message-Id: <20211003175743.3738710-4-f4bug@amsat.org>

target/mips: Use tcg_constant_i32() in gen_msa_elm_df()

Data Format is a 2-bit constant value.
Avoid using a TCG temporary by moving it to the constant pool.

Signed-off-by: Philippe Mathieu-Daudé <f4bug@amsat.org>
Reviewed-by: Richard Henderson <richard.henderson@linaro.org>
Message-Id: <20211003175743.3738710-3-f4bug@amsat.org>

target/mips: Remove unused register from MSA 2R/2RF instruction format

Commits cbe50b9a8e7 ("target-mips: add MSA VEC/2R format instructions")
and 3bdeb68866e ("target-mips: add MSA 2RF format instructions") added
the MSA 2R/2RF instructions. However these instructions don't use any
target vector register, so remove the unused TCG temporaries.

Reviewed-by: Richard Henderson <richard.henderson@linaro.org>
Signed-off-by: Philippe Mathieu-Daudé <f4bug@amsat.org>
Message-Id: <20211003175743.3738710-2-f4bug@amsat.org>

hw/mips/boston: Add FDT generator

Generate FDT on our own if no dtb argument supplied.
Avoid introducing unused device in FDT with user supplied dtb.

Signed-off-by: Jiaxun Yang <jiaxun.yang@flygoat.com>
[PMD: Fix coding style]
Signed-off-by: Philippe Mathieu-Daudé <f4bug@amsat.org>
Message-Id: <20211002184539.169-4-jiaxun.yang@flygoat.com>

hw/mips/boston: Allow loading elf kernel and dtb

ELF kernel allows us debugging much easier with DWARF symbols.

Signed-off-by: Jiaxun Yang <jiaxun.yang@flygoat.com>
Reviewed-by: Philippe Mathieu-Daudé <f4bug@amsat.org>
[PMD: Fix coding style]
Signed-off-by: Philippe Mathieu-Daudé <f4bug@amsat.org>
Message-Id: <20211002184539.169-3-jiaxun.yang@flygoat.com>

hw/mips/boston: Massage memory map information

Use memmap array to uinfy address of memory map.
That would allow us reuse address information for FDT generation.

Signed-off-by: Jiaxun Yang <jiaxun.yang@flygoat.com>
Reviewed-by: Philippe Mathieu-Daudé <f4bug@amsat.org>
[PMD: Use local 'regaddr' in gen_firmware(), fix coding style]
Signed-off-by: Philippe Mathieu-Daudé <f4bug@amsat.org>
Message-Id: <20211002184539.169-2-jiaxun.yang@flygoat.com>

target/mips: Check nanoMIPS DSP MULT[U] accumulator with Release 6

Per the "MIPS Architecture Extension: nanoMIPS32 DSP TRM" rev 0.04,
MULT and MULTU opcodes:

  The value of ac selects an accumulator numbered from 0 to 3.
  When ac=0, this refers to the original HI/LO register pair of the
  MIPS32 architecture.

  In Release 6 of the MIPS Architecture, accumulators are eliminated
  from MIPS32.

Ensure pre-Release 6 is restricted to HI/LO registers pair.

Fixes: 8b3698b2947 ("target/mips: Add emulation of DSP ASE for nanoMIPS - part 4")
Reviewed-by: Richard Henderson <richard.henderson@linaro.org>
Signed-off-by: Philippe Mathieu-Daudé <f4bug@amsat.org>

Merge remote-tracking branch 'remotes/rth/tags/pull-tcg-20211016' into staging

Move gdb singlestep to generic code
Fix cpu_common_props

# gpg: Signature made Sat 16 Oct 2021 11:13:54 AM PDT
# gpg:                using RSA key 7A481E78868B4DB6A85A05C064DF38E8AF7E215F
# gpg:                issuer "richard.henderson@linaro.org"
# gpg: Good signature from "Richard Henderson <richard.henderson@linaro.org>" [ultimate]

* remotes/rth/tags/pull-tcg-20211016: (24 commits)
  Revert "cpu: Move cpu_common_props to hw/core/cpu.c"
  target/xtensa: Drop check for singlestep_enabled
  target/tricore: Drop check for singlestep_enabled
  target/sh4: Drop check for singlestep_enabled
  target/s390x: Drop check for singlestep_enabled
  target/rx: Drop checks for singlestep_enabled
  target/riscv: Remove exit_tb and lookup_and_goto_ptr
  target/riscv: Remove dead code after exception
  target/ppc: Drop exit checks for singlestep_enabled
  target/openrisc: Drop checks for singlestep_enabled
  target/mips: Drop exit checks for singlestep_enabled
  target/mips: Fix single stepping
  target/microblaze: Drop checks for singlestep_enabled
  target/microblaze: Check CF_NO_GOTO_TB for DISAS_JUMP
  target/m68k: Drop checks for singlestep_enabled
  target/i386: Drop check for singlestep_enabled
  target/i386: Check CF_NO_GOTO_TB for dc->jmp_opt
  target/hppa: Drop checks for singlestep_enabled
  target/arm: Drop checks for singlestep_enabled
  target/hexagon: Drop checks for singlestep_enabled
  ...

Signed-off-by: Richard Henderson <richard.henderson@linaro.org>

Revert "cpu: Move cpu_common_props to hw/core/cpu.c"

This reverts commit 1b36e4f5a5de585210ea95f2257839c2312be28f.

Despite a comment saying why cpu_common_props cannot be placed in
a file that is compiled once, it was moved anyway.  Revert that.

Since then, Property is not defined in hw/core/cpu.h, so it is now
easier to declare a function to install the properties rather than
the Property array itself.

Cc: Eduardo Habkost <ehabkost@redhat.com>
Suggested-by: Peter Maydell <peter.maydell@linaro.org>
Signed-off-by: Richard Henderson <richard.henderson@linaro.org>

target/xtensa: Drop check for singlestep_enabled

GDB single-stepping is now handled generically.

Signed-off-by: Richard Henderson <richard.henderson@linaro.org>

target/tricore: Drop check for singlestep_enabled

GDB single-stepping is now handled generically.

Reviewed-by: Philippe Mathieu-Daudé <f4bug@amsat.org>
Signed-off-by: Richard Henderson <richard.henderson@linaro.org>

target/sh4: Drop check for singlestep_enabled

GDB single-stepping is now handled generically.

Reviewed-by: Philippe Mathieu-Daudé <f4bug@amsat.org>
Signed-off-by: Richard Henderson <richard.henderson@linaro.org>

target/s390x: Drop check for singlestep_enabled

GDB single-stepping is now handled generically.

Signed-off-by: Richard Henderson <richard.henderson@linaro.org>

target/rx: Drop checks for singlestep_enabled

GDB single-stepping is now handled generically.

Reviewed-by: Philippe Mathieu-Daudé <f4bug@amsat.org>
Signed-off-by: Richard Henderson <richard.henderson@linaro.org>

target/riscv: Remove exit_tb and lookup_and_goto_ptr

GDB single-stepping is now handled generically, which means
we don't need to do anything in the wrappers.

Reviewed-by: Alistair Francis <alistair.francis@wdc.com>
Signed-off-by: Richard Henderson <richard.henderson@linaro.org>

target/riscv: Remove dead code after exception

We have already set DISAS_NORETURN in generate_exception,
which makes the exit_tb unreachable.

Reviewed-by: Alistair Francis <alistair.francis@wdc.com>
Signed-off-by: Richard Henderson <richard.henderson@linaro.org>

target/ppc: Drop exit checks for singlestep_enabled

GDB single-stepping is now handled generically.
Reuse gen_debug_exception to handle architectural debug exceptions.

Signed-off-by: Richard Henderson <richard.henderson@linaro.org>

target/openrisc: Drop checks for singlestep_enabled

GDB single-stepping is now handled generically.

Reviewed-by: Philippe Mathieu-Daudé <f4bug@amsat.org>
Signed-off-by: Richard Henderson <richard.henderson@linaro.org>

target/mips: Drop exit checks for singlestep_enabled

GDB single-stepping is now handled generically.

Reviewed-by: Philippe Mathieu-Daudé <f4bug@amsat.org>
Signed-off-by: Richard Henderson <richard.henderson@linaro.org>

target/mips: Fix single stepping

As per an ancient comment in mips_tr_translate_insn about the
expectations of gdb, when restarting the insn in a delay slot
we also re-execute the branch.  Which means that we are
expected to execute two insns in this case.

This has been broken since 8b86d6d2580, where we forced max_insns
to 1 while single-stepping.  This resulted in an exit from the
translator loop after the branch but before the delay slot is
translated.

Increase the max_insns to 2 for this case.  In addition, bypass
the end-of-page check, for when the branch itself ends the page.

Reviewed-by: Philippe Mathieu-Daudé <f4bug@amsat.org>
Signed-off-by: Richard Henderson <richard.henderson@linaro.org>

target/microblaze: Drop checks for singlestep_enabled

GDB single-stepping is now handled generically.

Signed-off-by: Richard Henderson <richard.henderson@linaro.org>

target/microblaze: Check CF_NO_GOTO_TB for DISAS_JUMP

We were using singlestep_enabled as a proxy for whether
translator_use_goto_tb would always return false.

Signed-off-by: Richard Henderson <richard.henderson@linaro.org>

target/m68k: Drop checks for singlestep_enabled

GDB single-stepping is now handled generically.

Acked-by: Laurent Vivier <laurent@vivier.eu>
Signed-off-by: Richard Henderson <richard.henderson@linaro.org>

target/i386: Drop check for singlestep_enabled

GDB single-stepping is now handled generically.

Signed-off-by: Richard Henderson <richard.henderson@linaro.org>

target/i386: Check CF_NO_GOTO_TB for dc->jmp_opt

We were using singlestep_enabled as a proxy for whether
translator_use_goto_tb would always return false.

Signed-off-by: Richard Henderson <richard.henderson@linaro.org>

target/hppa: Drop checks for singlestep_enabled

GDB single-stepping is now handled generically.

Reviewed-by: Philippe Mathieu-Daudé <f4bug@amsat.org>
Signed-off-by: Richard Henderson <richard.henderson@linaro.org>

target/arm: Drop checks for singlestep_enabled

GDB single-stepping is now handled generically.

Signed-off-by: Richard Henderson <richard.henderson@linaro.org>

target/hexagon: Drop checks for singlestep_enabled

GDB single-stepping is now handled generically.

Reviewed-by: Philippe Mathieu-Daudé <f4bug@amsat.org>
Signed-off-by: Richard Henderson <richard.henderson@linaro.org>

target/cris: Drop checks for singlestep_enabled

GDB single-stepping is now handled generically.

Signed-off-by: Richard Henderson <richard.henderson@linaro.org>

target/avr: Drop checks for singlestep_enabled

GDB single-stepping is now handled generically.

Tested-by: Michael Rolnik <mrolnik@gmail.com>
Reviewed-by: Michael Rolnik <mrolnik@gmail.com>
Reviewed-by: Philippe Mathieu-Daudé <f4bug@amsat.org>
Signed-off-by: Richard Henderson <richard.henderson@linaro.org>

target/alpha: Drop checks for singlestep_enabled

GDB single-stepping is now handled generically.

Reviewed-by: Philippe Mathieu-Daudé <f4bug@amsat.org>
Signed-off-by: Richard Henderson <richard.henderson@linaro.org>

accel/tcg: Handle gdb singlestep in cpu_tb_exec

Currently the change in cpu_tb_exec is masked by the debug exception
being raised by the translators.  But this allows us to remove that code.

Signed-off-by: Richard Henderson <richard.henderson@linaro.org>

Merge remote-tracking branch 'remotes/ericb/tags/pull-nbd-2021-10-15' into staging

nbd patches for 2021-10-15

- Vladimir Sementsov-Ogievskiy: Consistent use of 64-bit parameters in
  block operations
- Hanna Reitz: Silence 32-bit compiler warning

# gpg: Signature made Fri 15 Oct 2021 02:08:10 PM PDT
# gpg:                using RSA key 71C2CC22B1C4602927D2F3AAA7A16B4A2527436A
# gpg: Good signature from "Eric Blake <eblake@redhat.com>" [full]
# gpg:                 aka "Eric Blake (Free Software Programmer) <ebb9@byu.net>" [full]
# gpg:                 aka "[jpeg image of size 6874]" [full]

* remotes/ericb/tags/pull-nbd-2021-10-15:
  block-backend: update blk_co_pwrite() and blk_co_pread() wrappers
  block-backend: fix blk_co_flush prototype to mention coroutine_fn
  block-backend: drop INT_MAX restriction from blk_check_byte_request()
  block-backend: blk_pread, blk_pwrite: rename count parameter to bytes
  block-backend: convert blk_aio_ functions to int64_t bytes paramter
  block-backend: convert blk_co_copy_range to int64_t bytes
  block-backend: convert blk_foo wrappers to use int64_t bytes parameter
  block-backend: drop blk_prw, use block-coroutine-wrapper
  block-coroutine-wrapper.py: support BlockBackend first argument
  block-backend: rename _do_ helper functions to _co_do_
  block-backend: convert blk_co_pdiscard to int64_t bytes
  block-backend: convert blk_co_pwritev_part to int64_t bytes
  block-backend: make blk_co_preadv() 64bit
  block-backend: blk_check_byte_request(): int64_t bytes
  qcow2: Silence clang -m32 compiler warning

Signed-off-by: Richard Henderson <richard.henderson@linaro.org>

block-backend: update blk_co_pwrite() and blk_co_pread() wrappers

Make bytes argument int64_t to be consistent with modern block-layer.
Callers should be OK with it as type becomes wider.

What is inside functions?

- Conversion from int64_t to size_t. Still, we
can't have a buffer larger than SIZE_MAX, therefore bytes should not be
larger than SIZE_MAX as well. Add an assertion.

- Passing to blk_co_pwritev() / blk_co_preadv() which already has
  int64_t bytes argument.

Signed-off-by: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com>
Message-Id: <20211007175243.642516-2-vsementsov@virtuozzo.com>
Reviewed-by: Eric Blake <eblake@redhat.com>
[eblake: spelling fix]
Signed-off-by: Eric Blake <eblake@redhat.com>

block-backend: fix blk_co_flush prototype to mention coroutine_fn

We already have this marker for the blk_co_flush function declaration in
block/block-backend.c. Add it in the header too.

Signed-off-by: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com>
Message-Id: <20211007175243.642516-1-vsementsov@virtuozzo.com>
Reviewed-by: Eric Blake <eblake@redhat.com>
[eblake: wording tweak]
Signed-off-by: Eric Blake <eblake@redhat.com>

block-backend: drop INT_MAX restriction from blk_check_byte_request()

blk_check_bytes_request is called from blk_co_do_preadv,
blk_co_do_pwritev_part, blk_co_do_pdiscard and blk_co_copy_range
before (maybe) calling throttle_group_co_io_limits_intercept() (which
has int64_t argument) and then calling corresponding bdrv_co_ function.
bdrv_co_ functions are OK with int64_t bytes as well.

So dropping the check for INT_MAX we just get same restrictions as in
bdrv_ layer: discard and write-zeroes goes through
bdrv_check_qiov_request() and are allowed to be 64bit. Other requests
go through bdrv_check_request32() and still restricted by INT_MAX
boundary.

Signed-off-by: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com>
Message-Id: <20211006131718.214235-13-vsementsov@virtuozzo.com>
Reviewed-by: Eric Blake <eblake@redhat.com>
Signed-off-by: Eric Blake <eblake@redhat.com>

block-backend: blk_pread, blk_pwrite: rename count parameter to bytes

To be consistent with declarations in include/sysemu/block-backend.h.

Signed-off-by: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com>
Message-Id: <20211006131718.214235-12-vsementsov@virtuozzo.com>
Reviewed-by: Eric Blake <eblake@redhat.com>
Signed-off-by: Eric Blake <eblake@redhat.com>

block-backend: convert blk_aio_ functions to int64_t bytes paramter

1. Convert bytes in BlkAioEmAIOCB:
  aio->bytes is only passed to already int64_t interfaces, and set in
  blk_aio_prwv, which is updated here.

2. For all updated functions the parameter type becomes wider so callers
   are safe.

3. In blk_aio_prwv we only store bytes to BlkAioEmAIOCB, which is
   updated here.

4. Other updated functions are wrappers on blk_aio_prwv.

Note that blk_aio_preadv and blk_aio_pwritev become safer: before this
commit, it's theoretically possible to pass qiov with size exceeding
INT_MAX, which than converted to int argument of blk_aio_prwv. Now it's
converted to int64_t which is a lot better. Still add assertions.

Signed-off-by: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com>
Message-Id: <20211006131718.214235-11-vsementsov@virtuozzo.com>
Reviewed-by: Eric Blake <eblake@redhat.com>
[eblake: tweak assertion and grammar]
Signed-off-by: Eric Blake <eblake@redhat.com>

block-backend: convert blk_co_copy_range to int64_t bytes

Function is updated so that parameter type becomes wider, so all
callers should be OK with it.

Look at blk_co_copy_range() itself: bytes is passed only to
blk_check_byte_request() and bdrv_co_copy_range(), which already have
int64_t bytes parameter, so we are OK.

Note that requests exceeding INT_MAX are still restricted by
blk_check_byte_request().

Signed-off-by: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com>
Message-Id: <20211006131718.214235-10-vsementsov@virtuozzo.com>
Reviewed-by: Eric Blake <eblake@redhat.com>
[eblake: grammar tweaks]
Signed-off-by: Eric Blake <eblake@redhat.com>

block-backend: convert blk_foo wrappers to use int64_t bytes parameter

Convert blk_pdiscard, blk_pwrite_compressed, blk_pwrite_zeroes.
These are just wrappers for functions with int64_t argument, so allow
passing int64_t as well. Parameter type becomes wider so all callers
should be OK with it.

Note that requests exceeding INT_MAX are still restricted by
blk_check_byte_request().

Note also that we don't (and are not going to) convert blk_pwrite and
blk_pread: these functions return number of bytes on success, so to
update them, we should change return type to int64_t as well, which
will lead to investigating and updating all callers which is too much.

So, blk_pread and blk_pwrite remain unchanged.

Signed-off-by: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com>
Message-Id: <20211006131718.214235-9-vsementsov@virtuozzo.com>
Reviewed-by: Eric Blake <eblake@redhat.com>
[eblake: grammar tweaks]
Signed-off-by: Eric Blake <eblake@redhat.com>

block-backend: drop blk_prw, use block-coroutine-wrapper

Let's drop hand-made coroutine wrappers and use coroutine wrapper
generation like in block/io.c.

Now, blk_foo() functions are written in same way as blk_co_foo() ones,
but wrap blk_do_foo() instead of blk_co_do_foo().

Signed-off-by: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com>
Message-Id: <20211006131718.214235-8-vsementsov@virtuozzo.com>
Reviewed-by: Eric Blake <eblake@redhat.com>
[eblake: spelling fix]
Signed-off-by: Eric Blake <eblake@redhat.com>

block-coroutine-wrapper.py: support BlockBackend first argument

Signed-off-by: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com>
Message-Id: <20211006131718.214235-7-vsementsov@virtuozzo.com>
Reviewed-by: Eric Blake <eblake@redhat.com>
Signed-off-by: Eric Blake <eblake@redhat.com>

block-backend: rename _do_ helper functions to _co_do_

This is a preparation to the following commit, to use automatic
coroutine wrapper generation.

Signed-off-by: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com>
Message-Id: <20211006131718.214235-6-vsementsov@virtuozzo.com>
Reviewed-by: Eric Blake <eblake@redhat.com>
Signed-off-by: Eric Blake <eblake@redhat.com>

block-backend: convert blk_co_pdiscard to int64_t bytes

We updated blk_do_pdiscard() and its wrapper blk_co_pdiscard(). Both
functions are updated so that the parameter type becomes wider, so all
callers should be OK with it.

Look at blk_do_pdiscard(): bytes is passed only to
blk_check_byte_request() and bdrv_co_pdiscard(), which already have
int64_t bytes parameter, so we are OK.

Note that requests exceeding INT_MAX are still restricted by
blk_check_byte_request().

Signed-off-by: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com>
Message-Id: <20211006131718.214235-5-vsementsov@virtuozzo.com>
Reviewed-by: Eric Blake <eblake@redhat.com>
[eblake: grammar tweaks]
Signed-off-by: Eric Blake <eblake@redhat.com>

block-backend: convert blk_co_pwritev_part to int64_t bytes

We convert blk_do_pwritev_part() and some wrappers:
blk_co_pwritev_part(), blk_co_pwritev(), blk_co_pwrite_zeroes().

All functions are converted so that the parameter type becomes wider, so
all callers should be OK with it.

Look at blk_do_pwritev_part() body:
bytes is passed to:

 - trace_blk_co_pwritev (we update it here)
 - blk_check_byte_request, throttle_group_co_io_limits_intercept,
   bdrv_co_pwritev_part - all already have int64_t argument.

Note that requests exceeding INT_MAX are still restricted by
blk_check_byte_request().

Signed-off-by: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com>
Message-Id: <20211006131718.214235-4-vsementsov@virtuozzo.com>
Reviewed-by: Eric Blake <eblake@redhat.com>
[eblake: grammar tweaks]
Signed-off-by: Eric Blake <eblake@redhat.com>

block-backend: make blk_co_preadv() 64bit

For both updated functions, the type of bytes becomes wider, so all callers
should be OK with it.

blk_co_preadv() only passes its arguments to blk_do_preadv().

blk_do_preadv() passes bytes to:

 - trace_blk_co_preadv, which is updated too
 - blk_check_byte_request, throttle_group_co_io_limits_intercept,
   bdrv_co_preadv, which are already int64_t.

Note that requests exceeding INT_MAX are still restricted by
blk_check_byte_request().

Signed-off-by: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com>
Message-Id: <20211006131718.214235-3-vsementsov@virtuozzo.com>
Reviewed-by: Eric Blake <eblake@redhat.com>
[eblake: grammar tweaks]
Signed-off-by: Eric Blake <eblake@redhat.com>

block-backend: blk_check_byte_request(): int64_t bytes

Rename size and make it int64_t to correspond to modern block layer,
which always uses int64_t for offset and bytes (not in blk layer yet,
which is a task for following commits).

All callers pass int or unsigned int.

So, for bytes in [0, INT_MAX] nothing is changed, for negative bytes we
now fail on "bytes < 0" check instead of "bytes > INT_MAX" check.

Note, that blk_check_byte_request() still doesn't allow requests
exceeding INT_MAX.

Signed-off-by: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com>
Message-Id: <20211006131718.214235-2-vsementsov@virtuozzo.com>
Reviewed-by: Eric Blake <eblake@redhat.com>
Signed-off-by: Eric Blake <eblake@redhat.com>

qcow2: Silence clang -m32 compiler warning

With -m32, size_t is generally only a uint32_t.  That makes clang
complain that in the assertion

  assert(qiov->size <= INT64_MAX);

the range of the type of qiov->size (size_t) is too small for any of its
values to ever exceed INT64_MAX.

Cast qiov->size to uint64_t to silence clang.

Fixes: f7ef38dd1310d7d9db76d0aa16899cbc5744f36d
       ("block: use int64_t instead of uint64_t in driver read
       handlers")
Signed-off-by: Hanna Reitz <hreitz@redhat.com>
Message-Id: <20211011155031.149158-1-hreitz@redhat.com>
Reviewed-by: Eric Blake <eblake@redhat.com>
Signed-off-by: Eric Blake <eblake@redhat.com>

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

qdev: Add JSON -device

- Add a JSON mode to the -device command line option
- net/vhost-{user,vdpa}: Fix device compatibility check
- Minor iotests fixes

# gpg: Signature made Fri 15 Oct 2021 07:41:22 AM PDT
# gpg:                using RSA key DC3DEB159A9AF95D3D7456FE7F09B272C88F2FD6
# gpg:                issuer "kwolf@redhat.com"
# gpg: Good signature from "Kevin Wolf <kwolf@redhat.com>" [full]

* remotes/kwolf/tags/for-upstream:
  vl: Enable JSON syntax for -device
  qdev: Base object creation on QDict rather than QemuOpts
  virtio-net: Avoid QemuOpts in failover_find_primary_device()
  virtio-net: Store failover primary opts pointer locally
  qdev: Add Error parameter to hide_device() callbacks
  qemu-option: Allow deleting opts during qemu_opts_foreach()
  softmmu/qdev-monitor: add error handling in qdev_set_id
  qdev: Make DeviceState.id independent of QemuOpts
  qdev: Avoid using string visitor for properties
  iotests/051: Fix typo
  iotests/245: Fix type for iothread property
  qom: Reduce use of error_propagate()
  net/vhost-vdpa: Fix device compatibility check
  net/vhost-user: Fix device compatibility check
  net: Introduce NetClientInfo.check_peer_type()

Signed-off-by: Richard Henderson <richard.henderson@linaro.org>

Merge remote-tracking branch 'remotes/thuth/tags/pull-request-2021-10-15' into staging

* Check kernel command line size on s390x
* Simplification of one of the SIGP instructions on s390x
* Cornelia stepping down as maintainer in some subsystems
* Update the dtc submodule to a proper release version

# gpg: Signature made Fri 15 Oct 2021 02:11:13 AM PDT
# gpg:                using RSA key 27B88847EEE0250118F3EAB92ED9D774FE702DB5
# gpg:                issuer "thuth@redhat.com"
# gpg: Good signature from "Thomas Huth <th.huth@gmx.de>" [full]
# gpg:                 aka "Thomas Huth <thuth@redhat.com>" [full]
# gpg:                 aka "Thomas Huth <th.huth@posteo.de>" [unknown]
# gpg:                 aka "Thomas Huth <huth@tuxfamily.org>" [full]

* remotes/thuth/tags/pull-request-2021-10-15:
  dtc: Update to version 1.6.1
  s390x virtio-ccw machine: step down as maintainer
  s390x/kvm: step down as maintainer
  vfio-ccw: step down as maintainer
  s390x: sigp: Force Set Architecture to return Invalid Parameter
  s390x/ipl: check kernel command line size

Signed-off-by: Richard Henderson <richard.henderson@linaro.org>

vl: Enable JSON syntax for -device

Like we already do for -object, introduce support for JSON syntax in
-device, which can be kept stable in the long term and guarantees that a
single code path with identical behaviour is used for both QMP and the
command line. Compared to the QemuOpts based code, the parser contains
less surprises and has support for non-scalar options (lists and
structs). Switching management tools to JSON means that we can more
easily change the "human" CLI syntax from QemuOpts to the keyval parser
later.

In the QAPI schema, a feature flag is added to the device-add command to
allow management tools to detect support for this.

Signed-off-by: Kevin Wolf <kwolf@redhat.com>
Reviewed-by: Eric Blake <eblake@redhat.com>
Message-Id: <20211008133442.141332-16-kwolf@redhat.com>
Reviewed-by: Michael S. Tsirkin <mst@redhat.com>
Tested-by: Peter Krempa <pkrempa@redhat.com>
Signed-off-by: Kevin Wolf <kwolf@redhat.com>

qdev: Base object creation on QDict rather than QemuOpts

QDicts are both what QMP natively uses and what the keyval parser
produces. Going through QemuOpts isn't useful for either one, so switch
the main device creation function to QDicts. By sharing more code with
the -object/object-add code path, we can even reduce the code size a
bit.

This commit doesn't remove the detour through QemuOpts from any code
path yet, but it allows the following commits to do so.

Signed-off-by: Kevin Wolf <kwolf@redhat.com>
Message-Id: <20211008133442.141332-15-kwolf@redhat.com>
Reviewed-by: Michael S. Tsirkin <mst@redhat.com>
Tested-by: Peter Krempa <pkrempa@redhat.com>
Signed-off-by: Kevin Wolf <kwolf@redhat.com>

virtio-net: Avoid QemuOpts in failover_find_primary_device()

Don't go through the global QemuOptsList, it is state of the legacy
command line parser and we will create devices that are not contained
in it. It is also just the command line configuration and not
necessarily the current runtime state.

Instead, look at the qdev device tree which has the current state of all
existing devices.

Signed-off-by: Kevin Wolf <kwolf@redhat.com>
Message-Id: <20211008133442.141332-14-kwolf@redhat.com>
Reviewed-by: Michael S. Tsirkin <mst@redhat.com>
Tested-by: Peter Krempa <pkrempa@redhat.com>
Signed-off-by: Kevin Wolf <kwolf@redhat.com>

virtio-net: Store failover primary opts pointer locally

Instead of accessing the global QemuOptsList, which really belong to the
command line parser and shouldn't be accessed from devices, store a
pointer to the QemuOpts in a new VirtIONet field.

This is not the final state, but just an intermediate step to get rid of
QemuOpts in devices. It will later be replaced with an options QDict.

Before this patch, two "primary" devices could be hidden for the same
standby device, but only one of them would actually be enabled and the
other one would be kept hidden forever, so this doesn't make sense.
After this patch, configuring a second primary device is an error.

Signed-off-by: Kevin Wolf <kwolf@redhat.com>
Message-Id: <20211008133442.141332-13-kwolf@redhat.com>
Reviewed-by: Michael S. Tsirkin <mst@redhat.com>
Tested-by: Peter Krempa <pkrempa@redhat.com>
Signed-off-by: Kevin Wolf <kwolf@redhat.com>

qdev: Add Error parameter to hide_device() callbacks

hide_device() is used for virtio-net failover, where the standby virtio
device delays creation of the primary device. It only makes sense to
have a single primary device for each standby device. Adding a second
one should result in an error instead of hiding it and never using it
afterwards.

Prepare for this by adding an Error parameter to the hide_device()
callback where virtio-net is informed about adding a primary device.

Signed-off-by: Kevin Wolf <kwolf@redhat.com>
Message-Id: <20211008133442.141332-12-kwolf@redhat.com>
Reviewed-by: Michael S. Tsirkin <mst@redhat.com>
Tested-by: Peter Krempa <pkrempa@redhat.com>
Signed-off-by: Kevin Wolf <kwolf@redhat.com>