Add support for probing RISC-V vector extension availability in
the backend. This information will be used when deciding whether
to use vector instructions in code generation.
Cache lg2(vlenb) for the backend. The storing of lg2(vlenb) means
we can convert all of the division into subtraction.
While the compiler doesn't support RISCV_HWPROBE_EXT_ZVE64X,
we use RISCV_HWPROBE_IMA_V instead. RISCV_HWPROBE_IMA_V is more
strictly constrainted than RISCV_HWPROBE_EXT_ZVE64X. At least in
current QEMU implemenation, the V vector extension depends on the
zve64d extension.
Signed-off-by: TANG Tiancheng <tangtiancheng.ttc@alibaba-inc.com>
Reviewed-by: Liu Zhiwei <zhiwei_liu@linux.alibaba.com>
Tested-by: Daniel Henrique Barboza <dbarboza@ventanamicro.com>
Message-ID: <20241007025700.47259-2-zhiwei_liu@linux.alibaba.com>
Signed-off-by: Richard Henderson <richard.henderson@linaro.org>
Move detection code out of tcg, similar to other hosts.
Reviewed-by: LIU Zhiwei <zhiwei_liu@linux.alibaba.com>
Reviewed-by: Alistair Francis <alistair.francis@wdc.com>
Reviewed-by: Philippe Mathieu-Daudé <philmd@linaro.org>
Reviewed-by: Daniel Henrique Barboza <dbarboza@ventanamicro.com>
Signed-off-by: Richard Henderson <richard.henderson@linaro.org>
This reverts commit b18236897c.
The x86-64 instruction set can now be tuned down to x86-64 v1
or i386 Pentium Pro.
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
This reverts commit 45ccdbcb24.
The x86-64 instruction set can now be tuned down to x86-64 v1
or i386 Pentium Pro.
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
Use inline assembly because no release compiler allows
per-function selection of the ISA.
Tested-by: Bibo Mao <maobibo@loongson.cn>
Signed-off-by: Richard Henderson <richard.henderson@linaro.org>
Split out host/bufferiszero.h.inc for x86, aarch64 and generic
in order to avoid an overlong ifdef ladder.
Reviewed-by: Philippe Mathieu-Daudé <philmd@linaro.org>
Signed-off-by: Richard Henderson <richard.henderson@linaro.org>
Reviewed-by: Song Gao <gaosong@loongson.cn>
Reviewed-by: Philippe Mathieu-Daudé <philmd@linaro.org>
Signed-off-by: Richard Henderson <richard.henderson@linaro.org>
QEMU now requires an x86-64-v2 host, which has the POPCNT instruction.
Use it freely in TCG-generated code.
Reviewed-by: Richard Henderson <richard.henderson@linaro.org>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
QEMU now requires an x86-64-v2 host, which has SSE2.
Use it freely in buffer_is_zero.
Reviewed-by: Richard Henderson <richard.henderson@linaro.org>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
QEMU now requires an x86-64-v2 host, which always has CMOV.
Use it freely in TCG generated code.
Reviewed-by: Richard Henderson <richard.henderson@linaro.org>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
The only user was the SSE4.1 variant of buffer_is_zero, which has
been removed; code to compute CPUINFO_SSE4 is dead.
Reviewed-by: Philippe Mathieu-Daudé <philmd@linaro.org>
Reviewed-by: Richard Henderson <richard.henderson@linaro.org>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
Since commit 139c1837db ("meson: rename included C source files
to .c.inc"), QEMU standard procedure for included C files is to
use *.c.inc.
Besides, since commit 6a0057aa22 ("docs/devel: make a statement
about includes") this is documented in the Coding Style:
If you do use template header files they should be named with
the ``.c.inc`` or ``.h.inc`` suffix to make it clear they are
being included for expansion.
Therefore rename 'store-insert-al16.h' as 'store-insert-al16.h.inc'
and 'load-extract-al16-al8.h' as 'load-extract-al16-al8.h.inc'.
Signed-off-by: Philippe Mathieu-Daudé <philmd@linaro.org>
Acked-by: Richard Henderson <richard.henderson@linaro.org>
Message-Id: <20240424173333.96148-3-philmd@linaro.org>
When compiling QEMU with Clang 17 on a s390x, the compilation fails:
In file included from ../accel/tcg/cputlb.c:32:
In file included from /root/qemu/include/exec/helper-proto-common.h:10:
In file included from /root/qemu/include/qemu/atomic128.h:62:
/root/qemu/host/include/generic/host/atomic128-ldst.h:68:15: error:
__sync builtin operation MUST have natural alignment (consider using __
atomic). [-Werror,-Wsync-alignment]
68 | } while (!__sync_bool_compare_and_swap_16(ptr_align, old, new.i));
| ^
In file included from ../accel/tcg/cputlb.c:32:
In file included from /root/qemu/include/exec/helper-proto-common.h:10:
In file included from /root/qemu/include/qemu/atomic128.h:61:
/root/qemu/host/include/generic/host/atomic128-cas.h:36:11: error:
__sync builtin operation MUST have natural alignment (consider using __a
tomic). [-Werror,-Wsync-alignment]
36 | r.i = __sync_val_compare_and_swap_16(ptr_align, c.i, n.i);
| ^
2 errors generated.
It's arguably a bug in Clang since we already use __builtin_assume_aligned()
to tell the compiler that the pointer is properly aligned. But according to
https://github.com/llvm/llvm-project/issues/69146 it seems like the Clang
folks don't see an easy fix on their side and recommend to use a type
declared with __attribute__((aligned(16))) to work around this problem.
Resolves: https://gitlab.com/qemu-project/qemu/-/issues/1934
Message-ID: <20231108085954.313071-1-thuth@redhat.com>
Reviewed-by: Philippe Mathieu-Daudé <philmd@linaro.org>
Reviewed-by: Richard Henderson <richard.henderson@linaro.org>
Signed-off-by: Thomas Huth <thuth@redhat.com>
While loongarch64 does not have a 128-bit cmpxchg, it does
have 128-bit atomic load and store via the vector unit.
Signed-off-by: Richard Henderson <richard.henderson@linaro.org>
Message-Id: <20230916220151.526140-6-richard.henderson@linaro.org>
Detect CRYPTO in cpuinfo; implement the accel hooks.
Reviewed-by: Daniel Henrique Barboza <danielhb413@gmail.com>
Signed-off-by: Richard Henderson <richard.henderson@linaro.org>
Add a primitive for InvSubBytes + InvShiftRows +
AddRoundKey + InvMixColumns.
Acked-by: Daniel P. Berrangé <berrange@redhat.com>
Reviewed-by: Philippe Mathieu-Daudé <philmd@linaro.org>
Signed-off-by: Richard Henderson <richard.henderson@linaro.org>
Add a primitive for InvSubBytes + InvShiftRows +
InvMixColumns + AddRoundKey.
Acked-by: Daniel P. Berrangé <berrange@redhat.com>
Signed-off-by: Richard Henderson <richard.henderson@linaro.org>
Add a primitive for SubBytes + ShiftRows + MixColumns + AddRoundKey.
Acked-by: Daniel P. Berrangé <berrange@redhat.com>
Signed-off-by: Richard Henderson <richard.henderson@linaro.org>
Add a primitive for InvSubBytes + InvShiftRows + AddRoundKey.
Reviewed-by: Daniel P. Berrangé <berrange@redhat.com>
Signed-off-by: Richard Henderson <richard.henderson@linaro.org>
Start adding infrastructure for accelerating guest AES.
Begin with a SubBytes + ShiftRows + AddRoundKey primitive.
Acked-by: Daniel P. Berrangé <berrange@redhat.com>
Signed-off-by: Richard Henderson <richard.henderson@linaro.org>
Move the code from tcg/. Fix a bug in that PPC_FEATURE2_ARCH_3_10
is actually spelled PPC_FEATURE2_ARCH_3_1.
Reviewed-by: Philippe Mathieu-Daudé <philmd@linaro.org>
Reviewed-by: Daniel Henrique Barboza <danielhb413@gmail.com>
Signed-off-by: Richard Henderson <richard.henderson@linaro.org>
The macOS catalina compiler produces an error for __int128_t
as the type for allocation with SSE inline asm constraint.
Create a new X86Int128Union type and use the vector type for
all SSE register inputs and outputs.
Tested-by: Peter Maydell <peter.maydell@linaro.org>
Signed-off-by: Richard Henderson <richard.henderson@linaro.org>
With CPUINFO_ATOMIC_VMOVDQA, we can perform proper atomic
load/store without cmpxchg16b.
Reviewed-by: Alex Bennée <alex.bennee@linaro.org>
Signed-off-by: Richard Henderson <richard.henderson@linaro.org>
With FEAT_LSE2, load and store of int128 is directly supported.
Reviewed-by: Alex Bennée <alex.bennee@linaro.org>
Signed-off-by: Richard Henderson <richard.henderson@linaro.org>
Use __sync_bool_compare_and_swap_16 to control the loop,
rather than a separate comparison.
Reviewed-by: Alex Bennée <alex.bennee@linaro.org>
Signed-off-by: Richard Henderson <richard.henderson@linaro.org>
Create both atomic16_read_ro and atomic16_read_rw.
Previously we pretended that we had atomic16_read in system mode,
because we "know" that all ram is always writable to the host.
Now, expose read-only and read-write versions all of the time.
For aarch64, do not fall back to __atomic_read_16 even if
supported by the compiler, to work around a clang bug.
Reviewed-by: Alex Bennée <alex.bennee@linaro.org>
Signed-off-by: Richard Henderson <richard.henderson@linaro.org>
These symbols will shortly become dynamic runtime tests and
therefore not appropriate for the preprocessor. Use the
matching CONFIG_* symbols for that purpose.
Reviewed-by: Alex Bennée <alex.bennee@linaro.org>
Signed-off-by: Richard Henderson <richard.henderson@linaro.org>
Not only the routines in ldst_atomicity.c.inc need markup,
but also the ones in the headers.
Reviewed-by: Alex Bennée <alex.bennee@linaro.org>
Signed-off-by: Richard Henderson <richard.henderson@linaro.org>
Separates the aarch64-specific portion into its own file.
Reviewed-by: Alex Bennée <alex.bennee@linaro.org>
Signed-off-by: Richard Henderson <richard.henderson@linaro.org>
Separates the aarch64-specific portion into its own file.
Reviewed-by: Alex Bennée <alex.bennee@linaro.org>
Signed-off-by: Richard Henderson <richard.henderson@linaro.org>
Move the code from tcg/. The only use of these bits so far
is with respect to the atomicity of tcg operations.
Reviewed-by: Alex Bennée <alex.bennee@linaro.org>
Reviewed-by: Peter Maydell <peter.maydell@linaro.org>
Signed-off-by: Richard Henderson <richard.henderson@linaro.org>
