target/hppa: Manage PSW_X and PSW_B in translator

PSW_X is cleared after every instruction, and only set by RFI.
PSW_B is cleared after every non-branch, or branch not taken,
and only set by taken branches.  We can clear both bits with a
single store, at most once per TB.  Taken branches set PSW_B,
at most once per TB.

Reviewed-by: Helge Deller <deller@gmx.de>
Signed-off-by: Richard Henderson <richard.henderson@linaro.org>

target/hppa: Split PSW X and B into their own field

Generally, both of these bits are cleared at the end of each
instruction.  By separating these, we will be able to clear
both with a single insn, instead of 2 or 3.

Reviewed-by: Helge Deller <deller@gmx.de>
Reviewed-by: Philippe Mathieu-Daudé <philmd@linaro.org>
Signed-off-by: Richard Henderson <richard.henderson@linaro.org>

target/hppa: Improve hppa_cpu_dump_state

Print both raw IAQ_Front and IAQ_Back as well as the GVAs.
Print control registers in system mode.
Print floating point registers if CPU_DUMP_FPU.

Reviewed-by: Helge Deller <deller@gmx.de>
Signed-off-by: Richard Henderson <richard.henderson@linaro.org>

target/hppa: Do not mask in copy_iaoq_entry

As with loads and stores, code offsets are kept intact until the
full gva is formed.  In qemu, this is in cpu_get_tb_cpu_state.

Reviewed-by: Helge Deller <deller@gmx.de>
Signed-off-by: Richard Henderson <richard.henderson@linaro.org>

target/hppa: Store full iaoq_f and page offset of iaoq_b in TB

In preparation for CF_PCREL. store the iaoq_f in 3 parts: high
bits in cs_base, middle bits in pc, and low bits in priv.
For iaoq_b, set a bit for either of space or page differing,
else the page offset.

Install iaq entries before goto_tb. The change to not record
the full direct branch difference in TB means that we have to
store at least iaoq_b before goto_tb.  But since a later change
to enable CF_PCREL will require both iaoq_f and iaoq_b to be
updated before goto_tb, go ahead and update both fields now.

Reviewed-by: Helge Deller <deller@gmx.de>
Signed-off-by: Richard Henderson <richard.henderson@linaro.org>

linux-user/hppa: Force all code addresses to PRIV_USER

The kernel does this along the return path to user mode.

Reviewed-by: Helge Deller <deller@gmx.de>
Signed-off-by: Richard Henderson <richard.henderson@linaro.org>

target/hppa: Use delay_excp for conditional trap on overflow

Reviewed-by: Helge Deller <deller@gmx.de>
Signed-off-by: Richard Henderson <richard.henderson@linaro.org>

target/hppa: Use delay_excp for conditional traps

Reviewed-by: Helge Deller <deller@gmx.de>
Signed-off-by: Richard Henderson <richard.henderson@linaro.org>

target/hppa: Introduce DisasDelayException

Allow an exception to be emitted at the end of the TranslationBlock,
leaving only the conditional branch inline.  Use it for simple
exception instructions like break, which happen to be nullified.

Reviewed-by: Helge Deller <deller@gmx.de>
Signed-off-by: Richard Henderson <richard.henderson@linaro.org>

target/hppa: Remove cond_free

Now that we do not need to free tcg temporaries, the only
thing cond_free does is reset the condition to never.
Instead, simply write a new condition over the old, which
may be simply cond_make_f() for the never condition.

The do_*_cond functions do the right thing with c or cf == 0,
so there's no need for a special case anymore.

Reviewed-by: Helge Deller <deller@gmx.de>
Reviewed-by: Philippe Mathieu-Daudé <philmd@linaro.org>
Signed-off-by: Richard Henderson <richard.henderson@linaro.org>

target/hppa: Use TCG_COND_TST* in trans_ftest

Reviewed-by: Helge Deller <deller@gmx.de>
Reviewed-by: Philippe Mathieu-Daudé <philmd@linaro.org>
Signed-off-by: Richard Henderson <richard.henderson@linaro.org>

target/hppa: Use registerfields.h for FPSR

Define all of the context dependent field definitions.
Use FIELD_EX32 and FIELD_DP32 with named fields instead
of extract32 and deposit32 with raw constants.

Reviewed-by: Helge Deller <deller@gmx.de>
Signed-off-by: Richard Henderson <richard.henderson@linaro.org>

target/hppa: Use TCG_COND_TST* in trans_bb_imm

Reviewed-by: Helge Deller <deller@gmx.de>
Reviewed-by: Philippe Mathieu-Daudé <philmd@linaro.org>
Signed-off-by: Richard Henderson <richard.henderson@linaro.org>

target/hppa: Use TCG_COND_TST* in do_unit_addsub

Reviewed-by: Helge Deller <deller@gmx.de>
Reviewed-by: Philippe Mathieu-Daudé <philmd@linaro.org>
Signed-off-by: Richard Henderson <richard.henderson@linaro.org>

target/hppa: Use TCG_COND_TST* in do_unit_zero_cond

Reviewed-by: Helge Deller <deller@gmx.de>
Signed-off-by: Richard Henderson <richard.henderson@linaro.org>

target/hppa: Use TCG_COND_TST* in do_log_cond

We can directly test bits of a 32-bit comparison without
zero or sign-extending an intermediate result.
We can directly test bit 0 for odd/even.

Reviewed-by: Helge Deller <deller@gmx.de>
Signed-off-by: Richard Henderson <richard.henderson@linaro.org>

target/hppa: Use TCG_COND_TST* in do_cond

We can directly test bits of a 32-bit comparison without
zero or sign-extending an intermediate result.
We can directly test bit 0 for odd/even.

Reviewed-by: Helge Deller <deller@gmx.de>
Signed-off-by: Richard Henderson <richard.henderson@linaro.org>

target/hppa: Rename cond_make_* helpers

Use 'v' for a variable that needs copying, 't' for a temp that
doesn't need copying, and 'i' for an immediate, and use this
naming for both arguments of the comparison.  So:

   cond_make_tmp -> cond_make_tt
   cond_make_0_tmp -> cond_make_ti
   cond_make_0 -> cond_make_vi
   cond_make -> cond_make_vv

Pass 0 explictly, rather than implicitly in the function name.

Reviewed-by: Helge Deller <deller@gmx.de>
Reviewed-by: Philippe Mathieu-Daudé <philmd@linaro.org>
Signed-off-by: Richard Henderson <richard.henderson@linaro.org>

target/hppa: Use displacements in DisasIAQE

This is a first step in enabling CF_PCREL, but for now
we regenerate the absolute address before writeback.

Reviewed-by: Helge Deller <deller@gmx.de>
Signed-off-by: Richard Henderson <richard.henderson@linaro.org>

target/hppa: Introduce and use DisasIAQE for branch management

Wrap offset and space together in one structure, ensuring
that they're copied together as required.

Reviewed-by: Helge Deller <deller@gmx.de>
Signed-off-by: Richard Henderson <richard.henderson@linaro.org>

target/hppa: Always make a copy in do_ibranch_priv

This simplifies callers, which might otherwise have
to make another copy.

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

target/hppa: Use umax in do_ibranch_priv

Using umax is clearer than the same operation using movcond.

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

target/hppa: Add space argument to do_ibranch

This allows unification of BE, BLR, BV, BVE with a common helper.
Since we can now track space with IAQ_Next, we can now let the
TranslationBlock continue across the delay slot with BE, BVE.

Reviewed-by: Helge Deller <deller@gmx.de>
Signed-off-by: Richard Henderson <richard.henderson@linaro.org>

target/hppa: Add space arguments to install_iaq_entries

Move space assighments to a central location.

Reviewed-by: Helge Deller <deller@gmx.de>
Signed-off-by: Richard Henderson <richard.henderson@linaro.org>

target/hppa: Add IASQ entries to DisasContext

Add variable to track space changes to IAQ.  So far, no such changes
are introduced, but the new checks vs ctx->iasq_b may eliminate an
unnecessary copy to cpu_iasq_f with e.g. BLR.

Reviewed-by: Helge Deller <deller@gmx.de>
Signed-off-by: Richard Henderson <richard.henderson@linaro.org>

target/hppa: Simplify TB end

Minimize the amount of code in hppa_tr_translate_insn advancing the
insn queue for the next insn.  Move the goto_tb path to hppa_tr_tb_stop.

Reviewed-by: Helge Deller <deller@gmx.de>
Signed-off-by: Richard Henderson <richard.henderson@linaro.org>

target/hppa: Skip nullified insns in unconditional dbranch path

Reviewed-by: Helge Deller <deller@gmx.de>
Signed-off-by: Richard Henderson <richard.henderson@linaro.org>

target/hppa: Delay computation of IAQ_Next

We no longer have to allocate a temp and perform an
addition before translation of the rest of the insn.

Reviewed-by: Helge Deller <deller@gmx.de>
Signed-off-by: Richard Henderson <richard.henderson@linaro.org>

target/hppa: Add install_link

Add a common routine for writing the return address.

Reviewed-by: Helge Deller <deller@gmx.de>
Signed-off-by: Richard Henderson <richard.henderson@linaro.org>

target/hppa: Add install_iaq_entries

Instead of two separate cpu_iaoq_entry calls, use one call to update
both IAQ_Front and IAQ_Back.  Simplify with an argument combination
that automatically handles a simple increment from Front to Back.

Reviewed-by: Helge Deller <deller@gmx.de>
Signed-off-by: Richard Henderson <richard.henderson@linaro.org>

target/hppa: Use CF_BP_PAGE instead of cpu_breakpoint_test

The generic tcg driver will have already checked for breakpoints.

Reviewed-by: Helge Deller <deller@gmx.de>
Signed-off-by: Richard Henderson <richard.henderson@linaro.org>

target/hppa: Allow prior nullification in do_ibranch

Simplify the function by not attempting a conditional move
on the branch destination -- just use nullify_over normally.

Reviewed-by: Helge Deller <deller@gmx.de>
Signed-off-by: Richard Henderson <richard.henderson@linaro.org>

target/hppa: Pass displacement to do_dbranch

Pass a displacement instead of an absolute value.

In trans_be, remove the user-only do_dbranch case.  The branch we are
attempting to optimize is to the zero page, which is perforce on a
different page than the code currently executing, which means that
we will *not* use a goto_tb.  Use a plain indirect branch instead,
which is what we got out of the attempted direct branch anyway.

Reviewed-by: Helge Deller <deller@gmx.de>
Reviewed-by: Philippe Mathieu-Daudé <philmd@linaro.org>
Signed-off-by: Richard Henderson <richard.henderson@linaro.org>

target/hppa: Move constant destination check into use_goto_tb

Share this check between gen_goto_tb and hppa_tr_translate_insn.

Reviewed-by: Helge Deller <deller@gmx.de>
Signed-off-by: Richard Henderson <richard.henderson@linaro.org>

target/hppa: Use hppa_form_gva_psw in hppa_cpu_get_pc

This function is for log_pc(), which needs to produce a
similar result to cpu_get_tb_cpu_state().

Reviewed-by: Helge Deller <deller@gmx.de>
Signed-off-by: Richard Henderson <richard.henderson@linaro.org>

target/hppa: Move cpu_get_tb_cpu_state out of line

Reviewed-by: Helge Deller <deller@gmx.de>
Signed-off-by: Richard Henderson <richard.henderson@linaro.org>

tcg/loongarch64: Fill out tcg_out_{ld,st} for vector regs

TCG register spill/fill uses tcg_out_ld/st with all types,
not necessarily going through INDEX_op_{ld,st}_vec.

Cc: qemu-stable@nongnu.org
Fixes: 16288ded944 ("tcg/loongarch64: Lower basic tcg vec ops to LSX")
Resolves: https://gitlab.com/qemu-project/qemu/-/issues/2336
Signed-off-by: Richard Henderson <richard.henderson@linaro.org>
Reviewed-by: Song Gao <gaosong@loongson.cn>
Tested-by: Song Gao <gaosong@loongson.cn>

accel/tcg: Remove cpu_ldsb_code / cpu_ldsw_code

Previous commits replaced them by translator_ld* calls.

Signed-off-by: Philippe Mathieu-Daudé <philmd@linaro.org>
Message-Id: <20240405131532.40913-1-philmd@linaro.org>
Signed-off-by: Richard Henderson <richard.henderson@linaro.org>

target/s390x: Use translator_lduw in get_next_pc

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

target/xtensa: Use translator_ldub in xtensa_insn_len

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

target/rx: Use translator_ld*

Reviewed-by: Yoshinori Sato <ysato@users.sourceforge.jp>
Reviewed-by: Philippe Mathieu-Daudé <philmd@linaro.org>
Signed-off-by: Richard Henderson <richard.henderson@linaro.org>

target/riscv: Use translator_ld* for everything

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

target/cris: Use cris_fetch in translate_v10.c.inc

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

target/cris: Use translator_ld* in cris_fetch

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

target/avr: Use translator_lduw

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

target/i386: Use translator_ldub for everything

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

target/microblaze: Use translator_ldl

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

target/hexagon: Use translator_ldl in pkt_crosses_page

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

target/s390x: Disassemble EXECUTEd instructions

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

target/s390x: Fix translator_fake_ld length

The ilen value extracted from ex_value is the length of the
EXECUTE instruction itself, and so is the increment to the pc.
However, the length of the synthetic insn is located in the
opcode like all other instructions.

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

accel/tcg: Introduce translator_fake_ld

Replace translator_fake_ldb, which required multiple calls,
with translator_fake_ld, which can take all data at once.

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

disas: Use translator_st to get disassembly data

Read from already translated pages, or saved mmio data.

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

disas: Split disas.c

The routines in disas-common.c are also used from disas-mon.c.
Otherwise the rest of disassembly is only used from tcg.
While we're at it, put host and target code into separate files.

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

accel/tcg: Return bool from TranslatorOps.disas_log

We have eliminated most uses of this hook.  Reduce
further by allowing the hook to handle only the
special cases, returning false for normal processing.

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

accel/tcg: Provide default implementation of disas_log

Almost all of the disas_log implementations are identical.
Unify them within translator_loop.

Drop extra Priv/Virt logging from target/riscv.

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

plugins: Merge  alloc_tcg_plugin_context into plugin_gen_tb_start

We don't need to allocate plugin context at startup,
we can wait until we actually use it.

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

plugins: Use DisasContextBase for qemu_plugin_tb_vaddr

We do not need to separately record the start of the TB.

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

plugins: Use DisasContextBase for qemu_plugin_insn_haddr

We can delay the computation of haddr until the plugin
actually requests it.

Reviewed-by: Pierrick Bouvier <pierrick.bouvier@linaro.org>
Signed-off-by: Richard Henderson <richard.henderson@linaro.org>

plugins: Read mem_only directly from TB cflags

Do not pass around a boolean between multiple structures,
just read it from the TranslationBlock in the TCGContext.

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

plugins: Use translator_st for qemu_plugin_insn_data

Use the bytes that we record for the entire TB, rather than
a per-insn GByteArray.  Record the length of the insn in
plugin_gen_insn_end rather than infering from the length
of the array.

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

accel/tcg: Implement translator_st

Copy data out of a completed translation.  This will be used
for both plugins and disassembly.

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

plugins: Copy memory in qemu_plugin_insn_data

Instead of returning a host pointer, copy the data into
storage provided by the caller.

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

accel/tcg: Record DisasContextBase in tcg_ctx for plugins

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

accel/tcg: Record when translator_fake_ldb is used

Remove left-over comment from commit dcd092a063
("accel/tcg: Improve can_do_io management").

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

accel/tcg: Record mmio bytes during translation

This will be able to replace plugin_insn_append, and will
be usable for disassembly.

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

accel/tcg: Cap the translation block when we encounter mmio

Do not allow translation to proceed beyond one insn with mmio,
as we will not be caching the TranslationBlock.

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

accel/tcg: Reorg translator_ld*

Reorg translator_access into translator_ld, with a more
memcpy-ish interface.  If both pages are in ram, do not
go through the caller's slow path.

Assert that the access is within the two pages that we are
prepared to protect, per TranslationBlock.  Allow access
prior to pc_first, so long as it is within the first page.

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

accel/tcg: Pass DisasContextBase to translator_fake_ldb

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

accel/tcg: Hide in_same_page outside of a target-specific context

While there are other methods that could be used to replace
TARGET_PAGE_MASK, the function is not really required outside
the context of target-specific translation.

This makes the header usable by target independent code.

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

accel/tcg: Use vaddr in translator_ld*

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

Merge tag 'ui-pull-request' of https://gitlab.com/marcandre.lureau/qemu into staging

UI: small fixes and improvements

# -----BEGIN PGP SIGNATURE-----
#
# iQJQBAABCAA6FiEEh6m9kz+HxgbSdvYt2ujhCXWWnOUFAmZDZEAcHG1hcmNhbmRy
# ZS5sdXJlYXVAcmVkaGF0LmNvbQAKCRDa6OEJdZac5UxvD/9HWbB8JdbV8lNCLePT
# a6RUWSqLyP/cV0FCw9URYgAjAYROO966dZopCH7+Sz6goC8tk3IFUoqL0LbtZQjK
# zMNueGRbwJj0iGMxFG4wuWBpBF6Dzc4sh90TF3XWSE8PMpWsDY+sP3VRu4sP1qu7
# OmCGTuSwNUugxazPLxvbTpLMnco9b+asAGlAU6WqpcURmia7XN7dBLGzfQ9vMxuc
# L5od+pPGfcxuj3ETMG+5OQlIZH1lmX3465LajkUDVxffNfznqMVDYyo4sKNW5KOY
# u420AoACeVsANWce1Aw2ekj1ETsvqxj23RClNIgdpDbMsGk9eM6eS+6vRctcM6z4
# wMH6GAKKI3AWj7Q6qY4096bcdNmYD/GOs9dgswqYjf+JLzEVcI1dHQ36K124nKH0
# t+9t3UUx1NBMwAp+EEN94W1ClwOZ0zvapS8zNaf76KIi9Eb4vrIyOlzdTM7SU4kC
# CQ4Tu9MBB5WIqzhsVtIH36zDBasgAU8DCtpelDY1AJiODGiQbfZi4yo8eEiQMS1s
# onixsXa7zyCCpmxwkYmvF54RbZFlPXxmdvu0jYxKddbEuTGX/Y3qvDAWv1kz6iJS
# iGmYtokfkv86XBCTGTAb3QEmFfOWcLnPc59Gg0TF3zyzY3q05nU/qjuIlgYedR/o
# TnnNYbyqXumojRCd69Dyy3THEg==
# =SW9v
# -----END PGP SIGNATURE-----
# gpg: Signature made Tue 14 May 2024 03:16:48 PM CEST
# gpg:                using RSA key 87A9BD933F87C606D276F62DDAE8E10975969CE5
# gpg:                issuer "marcandre.lureau@redhat.com"
# gpg: Good signature from "Marc-André Lureau <marcandre.lureau@redhat.com>" [full]
# gpg:                 aka "Marc-André Lureau <marcandre.lureau@gmail.com>" [full]

* tag 'ui-pull-request' of https://gitlab.com/marcandre.lureau/qemu:
  ui/sdl2: Allow host to power down screen
  ui/gtk: Fix mouse/motion event scaling issue with GTK display backend
  ui/gtk: Add gd_motion_event trace event
  ui/console: move QemuDmaBuf struct def to dmabuf.c
  ui/console: Use qemu_dmabuf_new() and free() helpers instead
  ui/console: Use qemu_dmabuf_set_..() helpers instead
  ui/console: Use qemu_dmabuf_get_..() helpers instead
  ui/console: new dmabuf.h and dmabuf.c for QemuDmaBuf struct and helpers
  ui/gtk: Check if fence_fd is equal to or greater than 0
  ui/gtk: Draw guest frame at refresh cycle
  Allow UNIX socket option for VNC websocket

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

Merge tag 'pull-request-2024-05-14' of https://gitlab.com/thuth/qemu into staging

* Fix the "tsan-build" CI job on the shared gitlab CI runners
* Bump minimum glib version and use URI code from the newer glib
* Fix error message from "configure" when C compiler is not working

# -----BEGIN PGP SIGNATURE-----
#
# iQJFBAABCAAvFiEEJ7iIR+7gJQEY8+q5LtnXdP5wLbUFAmZDXcYRHHRodXRoQHJl
# ZGhhdC5jb20ACgkQLtnXdP5wLbVPJw//bK/NuMKOHlnwgowkQ/x41t8nc0jAR38+
# aMhJBTSB+9EOlPd+/y7+IeFlD9lS2JzoX/CWeBrNlKc6juWQahABJYcvscmdGiYr
# a/dUy9iZoqJyY220TMjCWYwORRtNqPDXaiUIR8hBZZBmW51xs1hRc3aazxPm6dOD
# cj1yFKwWGY5g72SkRNTMWi3qWX1tXNOh7sbuhWKkZ3eiRCllHb0RwrhA341ze4TI
# ckmlSA6stMjls4XNAIAKVdRKLPE1BsJ/UKxpnOEO3F640cbe69B0+z13wIBNfTOY
# Mk3zSjrdLY6thSY+2iOb2FLt3wC5QCBjyRluv+0kwdSnz6xsafEDWNx5VneZH+Iu
# ZQWLGvN4qUUBBqHKY8eWnrsij3ABXioHLK8eHj2JuHidcG15tku/1cwAJvy/8P/O
# iup0elZ3MXaAk6ce3dwYY4t6QecuzqX9cdJkTuRNlzysK1xKQdBiYTdeZikfUAoM
# InuFUh732yPXDSiZcG+uMXUTAJXHWASr7bvPydDx/gL1tYGYBqYepfPF2uWYfNwg
# VZRgsN6WVDBGPyXv8Z7eQ9lye5JoAGYrSDxZE87q8RwRV5holiYDxtf10zeLz3Wf
# RI5L/bb2eFSHzi3quzOC1uLflLqNKwq+9UZEjdLv2z8zuhwVwxbcDV9+ox6zA8zi
# dnVC3Yp/3ik=
# =VRHz
# -----END PGP SIGNATURE-----
# gpg: Signature made Tue 14 May 2024 02:49:10 PM CEST
# 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]

* tag 'pull-request-2024-05-14' of https://gitlab.com/thuth/qemu:
  util/uri: Remove the old URI parsing code
  block/ssh: Use URI parsing code from glib
  block/nfs: Use URI parsing code from glib
  block/nbd: Use URI parsing code from glib
  block/gluster: Use URI parsing code from glib
  Remove glib compatibility code that is not required anymore
  Bump minimum glib version to v2.66
  gitlab: use 'setarch -R' to workaround tsan bug
  gitlab: use $MAKE instead of 'make'
  dockerfiles: add 'MAKE' env variable to remaining containers
  configure: Fix error message when C compiler is not working

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

ui/sdl2: Allow host to power down screen

By default, SDL disables the screen saver which prevents the host from powering
down the screen even if the screen is locked. This results in draining the
battery needlessly when the host isn't connected to a wall charger. Fix that by
enabling the screen saver.

Signed-off-by: Bernhard Beschow <shentey@gmail.com>
Acked-by: Marc-André Lureau <marcandre.lureau@redhat.com>
Message-ID: <20240512095945.1879-1-shentey@gmail.com>

ui/gtk: Fix mouse/motion event scaling issue with GTK display backend

Remove gtk_widget_get_scale_factor() usage from the calculation of
the motion events in the GTK backend to make it work correctly on
environments that have `gtk_widget_get_scale_factor() != 1`.

This scale factor usage had been introduced in the commit f14aab420c and
at that time the window size was used for calculating the things and it
was working correctly. However, in the commit 2f31663ed4 the logic
switched to use the widget size instead of window size and because of
the change the usage of scale factor becomes invalid (since widgets use
`vc->gfx.scale_{x, y}` for scaling).

Tested on Crostini on ChromeOS (15823.51.0) with an external display.

Fixes: 2f31663ed4 ("ui/gtk: use widget size for cursor motion event")
Fixes: f14aab420c ("ui: fix incorrect pointer position on highdpi with
gtk")

Signed-off-by: hikalium <hikalium@hikalium.com>
Acked-by: Marc-André Lureau <marcandre.lureau@redhat.com>
Message-Id: <20240512111435.30121-3-hikalium@hikalium.com>

ui/gtk: Add gd_motion_event trace event

Add gd_motion_event trace event for making it easy to debug
gd_motion_event related issues.

Signed-off-by: hikalium <hikalium@hikalium.com>
Reviewed-by: Marc-André Lureau <marcandre.lureau@redhat.com>
Message-Id: <20240512111435.30121-2-hikalium@hikalium.com>

ui/console: move QemuDmaBuf struct def to dmabuf.c

To complete privatizing process of QemuDmaBuf, QemuDmaBuf struct def
is moved to dmabuf.c

Suggested-by: Marc-André Lureau <marcandre.lureau@redhat.com>
Reviewed-by: Marc-André Lureau <marcandre.lureau@redhat.com>
Cc: Philippe Mathieu-Daudé <philmd@linaro.org>
Cc: Daniel P. Berrangé <berrange@redhat.com>
Cc: Vivek Kasireddy <vivek.kasireddy@intel.com>
Signed-off-by: Dongwon Kim <dongwon.kim@intel.com>
Message-Id: <20240508175403.3399895-7-dongwon.kim@intel.com>

ui/console: Use qemu_dmabuf_new() and free() helpers instead

This commit introduces utility functions for the creation and deallocation
of QemuDmaBuf instances. Additionally, it updates all relevant sections
of the codebase to utilize these new utility functions.

v7: remove prefix, "dpy_gl_" from all helpers
    qemu_dmabuf_free() returns without doing anything if input is null
    (Daniel P. Berrangé <berrange@redhat.com>)
    call G_DEFINE_AUTOPTR_CLEANUP_FUNC for qemu_dmabuf_free()
    (Daniel P. Berrangé <berrange@redhat.com>)

v8: Introduction of helpers was removed as those were already added
    by the previous commit

v9: set dmabuf->allow_fences to 'true' when dmabuf is created in
    virtio_gpu_create_dmabuf()/virtio-gpu-udmabuf.c

    removed unnecessary spaces were accidently added in the patch,
    'ui/console: Use qemu_dmabuf_new() a...'

v11: Calling qemu_dmabuf_close was removed as closing dmabuf->fd will be
     done in qemu_dmabuf_free anyway.
     (Daniel P. Berrangé <berrange@redhat.com>)

v12: --- Calling qemu_dmabuf_close separately as qemu_dmabuf_free doesn't
         do it.

     --- 'dmabuf' is now allocated space so it should be freed at the end of
         dbus_scanout_texture

v13: --- Immediately free dmabuf after it is released to prevent possible
         leaking of the ptr
         (Marc-André Lureau <marcandre.lureau@redhat.com>)

     --- Use g_autoptr macro to define *dmabuf for auto clean up instead of
         calling qemu_dmabuf_free
         (Marc-André Lureau <marcandre.lureau@redhat.com>)

v14: --- (vhost-user-gpu) Change qemu_dmabuf_free back to g_clear_pointer
         as it was done because of some misunderstanding (v13).

     --- (vhost-user-gpu) g->dmabuf[m->scanout_id] needs to be set to NULL
         to prevent freed dmabuf to be accessed again in case if(fd==-1)break;
         happens (before new dmabuf is allocated). Otherwise, it would cause
         invalid memory access when the same function is executed. Also NULL
         check should be done before qemu_dmabuf_close (it asserts dmabuf!=NULL.).
         (Marc-André Lureau <marcandre.lureau@redhat.com>)

Suggested-by: Marc-André Lureau <marcandre.lureau@redhat.com>
Cc: Philippe Mathieu-Daudé <philmd@linaro.org>
Cc: Daniel P. Berrangé <berrange@redhat.com>
Cc: Vivek Kasireddy <vivek.kasireddy@intel.com>
Signed-off-by: Dongwon Kim <dongwon.kim@intel.com>
Message-Id: <20240508175403.3399895-6-dongwon.kim@intel.com>

ui/console: Use qemu_dmabuf_set_..() helpers instead

This commit updates all occurrences where these fields were
set directly have been updated to utilize helper functions.

v7: removed prefix, "dpy_gl_" from all helpers

v8: Introduction of helpers was removed as those were already added
    by the previous commit

Suggested-by: Marc-André Lureau <marcandre.lureau@redhat.com>
Reviewed-by: Marc-André Lureau <marcandre.lureau@redhat.com>
Cc: Philippe Mathieu-Daudé <philmd@linaro.org>
Cc: Daniel P. Berrangé <berrange@redhat.com>
Cc: Vivek Kasireddy <vivek.kasireddy@intel.com>
Signed-off-by: Dongwon Kim <dongwon.kim@intel.com>
Message-Id: <20240508175403.3399895-5-dongwon.kim@intel.com>

ui/console: Use qemu_dmabuf_get_..() helpers instead

This commit updates all instances where fields within the QemuDmaBuf
struct are directly accessed, replacing them with calls to these new
helper functions.

v6: fix typos in helper names in ui/spice-display.c

v7: removed prefix, "dpy_gl_" from all helpers

v8: Introduction of helpers was removed as those were already added
    by the previous commit

v11: -- Use new qemu_dmabuf_close() instead of close(qemu_dmabuf_get_fd()).
        (Daniel P. Berrangé <berrange@redhat.com>)
     -- Use new qemu_dmabuf_dup_fd() instead of dup(qemu_dmabuf_get_fd()).
        (Daniel P. Berrangé <berrange@redhat.com>)

Suggested-by: Marc-André Lureau <marcandre.lureau@redhat.com>
Reviewed-by: Marc-André Lureau <marcandre.lureau@redhat.com>
Cc: Philippe Mathieu-Daudé <philmd@linaro.org>
Cc: Daniel P. Berrangé <berrange@redhat.com>
Cc: Vivek Kasireddy <vivek.kasireddy@intel.com>
Signed-off-by: Dongwon Kim <dongwon.kim@intel.com>
Message-Id: <20240508175403.3399895-4-dongwon.kim@intel.com>

ui/console: new dmabuf.h and dmabuf.c for QemuDmaBuf struct and helpers

New header and source files are added for containing QemuDmaBuf struct
definition and newly introduced helpers for creating/freeing the struct
and accessing its data.

v10: Change the license type for both dmabuf.h and dmabuf.c from MIT to
     GPL to be in line with QEMU's default license

v11: -- Added new helpers, qemu_dmabuf_close for closing dmabuf->fd,
        qemu_dmabuf_dup_fd for duplicating dmabuf->fd
        (Daniel P. Berrangé <berrange@redhat.com>)

     -- Let qemu_dmabuf_fee to call qemu_dmabuf_close before freeing
        the struct to make sure fd is closed.
        (Daniel P. Berrangé <berrange@redhat.com>)

v12: Not closing fd in qemu_dmabuf_free because there are cases fd
     should still be available even after the struct is destroyed
     (e.g. virtio-gpu: res->dmabuf_fd).

Suggested-by: Marc-André Lureau <marcandre.lureau@redhat.com>
Reviewed-by: Marc-André Lureau <marcandre.lureau@redhat.com>
Cc: Philippe Mathieu-Daudé <philmd@linaro.org>
Cc: Daniel P. Berrangé <berrange@redhat.com>
Cc: Vivek Kasireddy <vivek.kasireddy@intel.com>
Signed-off-by: Dongwon Kim <dongwon.kim@intel.com>
Message-Id: <20240508175403.3399895-3-dongwon.kim@intel.com>

ui/gtk: Check if fence_fd is equal to or greater than 0

'fence_fd' needs to be validated always before being referenced
And the passing condition should include '== 0' as 0 is a valid
value for the file descriptor.

Suggested-by: Marc-André Lureau <marcandre.lureau@redhat.com>
Reviewed-by: Daniel P. Berrangé <berrange@redhat.com>
Cc: Philippe Mathieu-Daudé <philmd@linaro.org>
Cc: Daniel P. Berrangé <berrange@redhat.com>
Cc: Vivek Kasireddy <vivek.kasireddy@intel.com>
Signed-off-by: Dongwon Kim <dongwon.kim@intel.com>
Message-Id: <20240508175403.3399895-2-dongwon.kim@intel.com>

ui/gtk: Draw guest frame at refresh cycle

Draw routine needs to be manually invoked in the next refresh
if there is a scanout blob from the guest. This is to prevent
a situation where there is a scheduled draw event but it won't
happen bacause the window is currently in inactive state
(minimized or tabified). If draw is not done for a long time,
gl_block timeout and/or fence timeout (on the guest) will happen
eventually.

v2: Use gd_gl_area_draw(vc) in gtk-gl-area.c

Suggested-by: Vivek Kasireddy <vivek.kasireddy@intel.com>
Cc: Gerd Hoffmann <kraxel@redhat.com>
Cc: Marc-André Lureau <marcandre.lureau@redhat.com>
Cc: Daniel P. Berrangé <berrange@redhat.com>
Signed-off-by: Dongwon Kim <dongwon.kim@intel.com>
Acked-by: Marc-André Lureau <marcandre.lureau@redhat.com>
Message-Id: <20240426225059.3871283-1-dongwon.kim@intel.com>

Allow UNIX socket option for VNC websocket

- Remove unix socket option limitation for VNC websocket
- Reflect websocket option changes in documentation

Signed-off-by: Sergii Zasenko <sergii@zasenko.name>
Reviewed-by: Marc-André Lureau <marcandre.lureau@redhat.com>
Message-Id: <20230724100353.16628-1-sergii@zasenko.name>

util/uri: Remove the old URI parsing code

Now that we switched all consumers of the URI code to use the URI
parsing functions from glib instead, we can remove our internal
URI parsing code since it is not used anymore.

Reviewed-by: Eric Blake <eblake@redhat.com>
Message-ID: <20240418101056.302103-14-thuth@redhat.com>
Signed-off-by: Thomas Huth <thuth@redhat.com>

block/ssh: Use URI parsing code from glib

Since version 2.66, glib has useful URI parsing functions, too.
Use those instead of the QEMU-internal ones to be finally able
to get rid of the latter.

While we're at it, also emit a warning when encountering unknown
parameters in the URI, so that the users have a chance to detect
their typos or other mistakes.

Reviewed-by: Richard W.M. Jones <rjones@redhat.com>
Message-ID: <20240418101056.302103-13-thuth@redhat.com>
Signed-off-by: Thomas Huth <thuth@redhat.com>

block/nfs: Use URI parsing code from glib

Since version 2.66, glib has useful URI parsing functions, too.
Use those instead of the QEMU-internal ones to be finally able
to get rid of the latter.

While we're at it, slightly rephrase one of the error messages:
Use "Invalid value..." instead of "Illegal value..." since the
latter rather sounds like the users were breaking a law here.

Reviewed-by: Eric Blake <eblake@redhat.com>
Message-ID: <20240418101056.302103-12-thuth@redhat.com>
Signed-off-by: Thomas Huth <thuth@redhat.com>

block/nbd: Use URI parsing code from glib

Since version 2.66, glib has useful URI parsing functions, too.
Use those instead of the QEMU-internal ones to be finally able
to get rid of the latter. The g_uri_get_host() also takes care
of removing the square brackets from IPv6 addresses, so we can
drop that part of the QEMU code now, too.

Reviewed-by: Richard W.M. Jones <rjones@redhat.com>
Reviewed-by: Eric Blake <eblake@redhat.com>
Message-ID: <20240418101056.302103-11-thuth@redhat.com>
Signed-off-by: Thomas Huth <thuth@redhat.com>

block/gluster: Use URI parsing code from glib

Since version 2.66, glib has useful URI parsing functions, too.
Use those instead of the QEMU-internal ones to be finally able
to get rid of the latter.

Since g_uri_get_path() returns a const pointer, we also need to
tweak the parameter of parse_volume_options() (where we use the
result of g_uri_get_path() as input).

Reviewed-by: Eric Blake <eblake@redhat.com>
Reviewed-by: Daniel P. Berrangé <berrange@redhat.com>
Message-Id: <20240418101056.302103-10-thuth@redhat.com>
Signed-off-by: Thomas Huth <thuth@redhat.com>

Remove glib compatibility code that is not required anymore

Now that we bumped the minimum glib version to 2.66, we can drop
the old code.

Suggested-by: Paolo Bonzini <pbonzini@redhat.com>
Reviewed-by: Daniel P. Berrangé <berrange@redhat.com>
Message-ID: <20240418101056.302103-9-thuth@redhat.com>
Signed-off-by: Thomas Huth <thuth@redhat.com>

Bump minimum glib version to v2.66

Now that we dropped support for CentOS 8 and Ubuntu 20.04, we can
look into bumping the glib version to a new minimum for further
clean-ups. According to repology.org, available versions are:

 CentOS Stream 9:       2.66.7
 Debian 11:             2.66.8
 Fedora 38:             2.74.1
 Freebsd:               2.78.4
 Homebrew:              2.80.0
 Openbsd:               2.78.4
 OpenSuse leap 15.5:    2.70.5
 pkgsrc_current:        2.78.4
 Ubuntu 22.04:          2.72.1

Thus it should be safe to bump the minimum glib version to 2.66 now.
Version 2.66 comes with new functions for URI parsing which will
allow further clean-ups in the following patches.

Reviewed-by: Daniel P. Berrangé <berrange@redhat.com>
Message-ID: <20240418101056.302103-8-thuth@redhat.com>
Signed-off-by: Thomas Huth <thuth@redhat.com>

gitlab: use 'setarch -R' to workaround tsan bug

The TSAN job started failing when gitlab rolled out their latest
release. The root cause is a change in the Google COS version used
on shared runners. This brings a kernel running with

 vm.mmap_rnd_bits = 31

which is incompatible with TSAN in LLVM < 18, which only supports
upto '28'. LLVM 18 can support upto '30', and failing that will
re-exec itself to turn off VA randomization.

Our LLVM is too old for now, but we can run with 'setarch -R make ..'
to turn off VA randomization ourselves.

Signed-off-by: Daniel P. Berrangé <berrange@redhat.com>
Reviewed-by: Thomas Huth <thuth@redhat.com>
Message-ID: <20240513111551.488088-4-berrange@redhat.com>
Signed-off-by: Thomas Huth <thuth@redhat.com>

gitlab: use $MAKE instead of 'make'

The lcitool generated containers have '$MAKE' set to the path
of the right 'make' binary. Using the env variable makes it
possible to override the choice per job.

Signed-off-by: Daniel P. Berrangé <berrange@redhat.com>
Reviewed-by: Thomas Huth <thuth@redhat.com>
Reviewed-by: Philippe Mathieu-Daudé <philmd@linaro.org>
Message-ID: <20240513111551.488088-3-berrange@redhat.com>
Signed-off-by: Thomas Huth <thuth@redhat.com>

dockerfiles: add 'MAKE' env variable to remaining containers

All the lcitool generated containers define a "MAKE" env. It will be
convenient for later patches if all containers do this.

Signed-off-by: Daniel P. Berrangé <berrange@redhat.com>
Reviewed-by: Thomas Huth <thuth@redhat.com>
Reviewed-by: Philippe Mathieu-Daudé <philmd@linaro.org>
Message-ID: <20240513111551.488088-2-berrange@redhat.com>
Signed-off-by: Thomas Huth <thuth@redhat.com>

configure: Fix error message when C compiler is not working

If you try to run the configure script on a system without a working
C compiler, you get a very misleading error message:

 ERROR: Unrecognized host OS (uname -s reports 'Linux')

Some people already opened bug tickets because of this problem:

 https://gitlab.com/qemu-project/qemu/-/issues/2057
 https://gitlab.com/qemu-project/qemu/-/issues/2288

We should rather tell the user that we were not able to use the C
compiler instead, otherwise they will have a hard time to figure
out what was going wrong.

While we're at it, let's also suppress the "unrecognized host CPU"
message in this case since it is rather misleading than helpful.

Fixes: 264b803721 ("configure: remove compiler sanity check")
Message-ID: <20240513114010.51608-1-thuth@redhat.com>
Reviewed-by: Philippe Mathieu-Daudé <philmd@linaro.org>
Signed-off-by: Thomas Huth <thuth@redhat.com>

Merge tag 'for-upstream' of https://gitlab.com/bonzini/qemu into staging

* target/i386: miscellaneous changes, mostly TCG-related
* fix --without-default-devices build
* fix --without-default-devices qtests on s390x and arm

# -----BEGIN PGP SIGNATURE-----
#
# iQFIBAABCAAyFiEE8TM4V0tmI4mGbHaCv/vSX3jHroMFAmY+JWIUHHBib256aW5p
# QHJlZGhhdC5jb20ACgkQv/vSX3jHroOGmwf+JKY/i7ihXvfINQIRSKaz+H7KM3Br
# BGv/iXj4hrRA+zflcZswwoWmPrkrXM3J5JqGG6zTqqhGne+fRKt60KBFwn+lRaMY
# n48icR4zOSaEcGKBOFKs9CB1JgL7SWMe+fZ8d02amYlIZ005af0d69ACenF9r/oX
# pTxYIrR90FdZStbF4Yl0G5CzMLBdHZd/b6bMNmbefVPv3/d2zuL7VgqLX3y3J0ee
# ASYkYjn8Wpda4KX9s2rvH9ENXj80Q7EqhuDvoBlyK72/2lE5aTojbUiyGB4n5AuX
# 5OHA+0HEpuCXXToijOeDXD1NDOk9E5DP8cEwwZfZ2gjWKjja0U6OODGLVw==
# =woTe
# -----END PGP SIGNATURE-----
# gpg: Signature made Fri 10 May 2024 03:47:14 PM CEST
# gpg:                using RSA key F13338574B662389866C7682BFFBD25F78C7AE83
# gpg:                issuer "pbonzini@redhat.com"
# gpg: Good signature from "Paolo Bonzini <bonzini@gnu.org>" [full]
# gpg:                 aka "Paolo Bonzini <pbonzini@redhat.com>" [full]

* tag 'for-upstream' of https://gitlab.com/bonzini/qemu: (27 commits)
  configs: disable emulators that require it if libfdt is not found
  hw/xtensa: require libfdt
  kconfig: express dependency of individual boards on libfdt
  kconfig: allow compiling out QEMU device tree code per target
  meson: move libfdt together with other dependencies
  meson: pick libfdt from common_ss when building target-specific files
  tests/qtest: arm: fix operation in a build without any boards or devices
  i386: select correct components for no-board build
  hw/i386: move rtc-reset-reinjection command out of hw/rtc
  hw/i386: split x86.c in multiple parts
  i386: pc: remove unnecessary MachineClass overrides
  i386: correctly select code in hw/i386 that depends on other components
  xen: register legacy backends via xen_backend_init
  xen: initialize legacy backends from xen_bus_init()
  tests/qtest: s390x: fix operation in a build without any boards or devices
  s390x: select correct components for no-board build
  s390: move css_migration_enabled from machine to css.c
  s390_flic: add migration-enabled property
  s390x: move s390_cpu_addr2state to target/s390x/sigp.c
  sh4: select correct components for no-board build
  ...

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

configs: disable emulators that require it if libfdt is not found

Since boards can express their dependency on libfdt and
system/device_tree.c, only leave TARGET_NEED_FDT if the target has a
hard dependency.

Those emulators will be skipped if libfdt is disabled, or if it
is "auto" and not found and --disable-download is passed; unless
the target is mentioned explicitly in --target-list, in which case
the build will fail.

Reviewed-by: Philippe Mathieu-Daudé <philmd@linaro.org>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>

hw/xtensa: require libfdt

All other boards require libfdt if it can be used (including for example
i386/x86_64), so change the "imply" to "select" and always allow -dtb
in qemu-system-xtensa.

Reviewed-by: Philippe Mathieu-Daudé <philmd@linaro.org>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>

kconfig: express dependency of individual boards on libfdt

Now that boards are enabled by default and the "CONFIG_FOO=y"
entries are gone from configs/devices/, there cannot be any more
a conflicts between the default contents of configs/devices/
and a failed "depends on" clause.

With this change, each individual board or target can express
whether it needs FDT.  It can then include the common code in the
build via "select DEVICE_TREE", which will also as tell meson to link
with libfdt.

This allows building non-microvm x86 emulators without having
libfdt available.

Reviewed-by: Philippe Mathieu-Daudé <philmd@linaro.org>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>

kconfig: allow compiling out QEMU device tree code per target

Introduce a new Kconfig symbol, CONFIG_DEVICE_TREE, that specifies whether
to include the common device tree code in system/device_tree.c and to
link to libfdt.  For now, include it unconditionally if libfdt is
available.

Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>

meson: move libfdt together with other dependencies

Move the libfdt detection code together with other dependencies instead
of keeping it with subprojects.  This has the disadvantage of performing
the detection even if no target requires libfdt; but it has the advantage
that Kconfig will be able to observe the availability of the library.

Reviewed-by: Philippe Mathieu-Daudé <philmd@linaro.org>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>