Support all of the easy GM block sizes.
Use direct memory operations, since the pointers are aligned.
While BS=2 (16 bytes, 1 tag) is a legal setting, that requires
an atomic store of one nibble. This is not difficult, but there
is also no point in supporting it until required.
Note that cortex-a710 sets GM blocksize to match its cacheline
size of 64 bytes. I expect many implementations will also
match the cacheline, which makes 16 bytes very unlikely.
Signed-off-by: Richard Henderson <richard.henderson@linaro.org>
Reviewed-by: Peter Maydell <peter.maydell@linaro.org>
Message-id: 20230811214031.171020-4-richard.henderson@linaro.org
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
Currently we list all the Arm decodetree files together and add them
unconditionally to arm_ss. This means we build them for both
qemu-system-aarch64 and qemu-system-arm. However, some of them are
AArch64-specific, so there is no need to build them for
qemu-system-arm. (Meson is smart enough to notice that the generated
.c.inc file is not used by any objects that go into qemu-system-arm,
so we only unnecessarily run decodetree, not anything more
heavyweight like a recompile or relink, but it's still unnecessary
work.)
Split gen into gen_a32 and gen_a64, and only add gen_a64 for
TARGET_AARCH64 compiles.
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
Reviewed-by: Philippe Mathieu-Daudé <philmd@linaro.org>
Message-id: 20230718104628.1137734-1-peter.maydell@linaro.org
In commit 0b188ea05a we changed the implementation of
trans_CSEL() to use tcg_constant_i32(). However, this change
was incorrect, because the implementation of the function
sets up the TCGv_i32 rn and rm to be either zero or else
a TCG temp created in load_reg(), and these TCG temps are
then in both cases written to by the emitted TCG ops.
The result is that we hit a TCG assertion:
qemu-system-arm: ../../tcg/tcg.c:4455: tcg_reg_alloc_mov: Assertion `!temp_readonly(ots)' failed.
(or on a non-debug build, just produce a garbage result)
Adjust the code so that rn and rm are always writeable
temporaries whether the instruction is using the special
case "0" or a normal register as input.
Cc: qemu-stable@nongnu.org
Fixes: 0b188ea05a ("target/arm: Use tcg_constant in trans_CSEL")
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
Reviewed-by: Richard Henderson <richard.henderson@linaro.org>
Message-id: 20230727103906.2641264-1-peter.maydell@linaro.org
This implements the AESIMC instruction. We have converted everything
to crypto/aes-round.h; crypto/aes.h is no longer needed.
Reviewed-by: Philippe Mathieu-Daudé <philmd@linaro.org>
Signed-off-by: Richard Henderson <richard.henderson@linaro.org>
Split these helpers so that we are not passing 'decrypt'
within the simd descriptor.
Reviewed-by: Philippe Mathieu-Daudé <philmd@linaro.org>
Signed-off-by: Richard Henderson <richard.henderson@linaro.org>
We do not currently have a table in crypto/ for just MixColumns.
Move both tables for consistency.
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>
The lack of SVE memory instrumentation has been an omission in plugin
handling since it was introduced. Fortunately we can utilise the
probe_* functions to force all all memory access to follow the slow
path. We do this by checking the access type and presence of plugin
memory callbacks and if set return the TLB_MMIO flag.
We have to jump through a few hoops in user mode to re-use the flag
but it was the desired effect:
./qemu-system-aarch64 -display none -serial mon:stdio \
-M virt -cpu max -semihosting-config enable=on \
-kernel ./tests/tcg/aarch64-softmmu/memory-sve \
-plugin ./contrib/plugins/libexeclog.so,ifilter=st1w,afilter=0x40001808 -d plugin
gives (disas doesn't currently understand st1w):
0, 0x40001808, 0xe54342a0, ".byte 0xa0, 0x42, 0x43, 0xe5", store, 0x40213010, RAM, store, 0x40213014, RAM, store, 0x40213018, RAM
And for user-mode:
./qemu-aarch64 \
-plugin contrib/plugins/libexeclog.so,afilter=0x4007c0 \
-d plugin \
./tests/tcg/aarch64-linux-user/sha512-sve
gives:
1..10
ok 1 - do_test(&tests[i])
0, 0x4007c0, 0xa4004b80, ".byte 0x80, 0x4b, 0x00, 0xa4", load, 0x5500800370, load, 0x5500800371, load, 0x5500800372, load, 0x5500800373, load, 0x5500800374, load, 0x5500800375, load, 0x5500800376, load, 0x5500800377, load, 0x5500800378, load, 0x5500800379, load, 0x550080037a, load, 0x550080037b, load, 0x550080037c, load, 0x550080037d, load, 0x550080037e, load, 0x550080037f, load, 0x5500800380, load, 0x5500800381, load, 0x5500800382, load, 0x5500800383, load, 0x5500800384, load, 0x5500800385, load, 0x5500800386, lo
ad, 0x5500800387, load, 0x5500800388, load, 0x5500800389, load, 0x550080038a, load, 0x550080038b, load, 0x550080038c, load, 0x550080038d, load, 0x550080038e, load, 0x550080038f, load, 0x5500800390, load, 0x5500800391, load, 0x5500800392, load, 0x5500800393, load, 0x5500800394, load, 0x5500800395, load, 0x5500800396, load, 0x5500800397, load, 0x5500800398, load, 0x5500800399, load, 0x550080039a, load, 0x550080039b, load, 0x550080039c, load, 0x550080039d, load, 0x550080039e, load, 0x550080039f, load, 0x55008003a0, load, 0x55008003a1, load, 0x55008003a2, load, 0x55008003a3, load, 0x55008003a4, load, 0x55008003a5, load, 0x55008003a6, load, 0x55008003a7, load, 0x55008003a8, load, 0x55008003a9, load, 0x55008003aa, load, 0x55008003ab, load, 0x55008003ac, load, 0x55008003ad, load, 0x55008003ae, load, 0x55008003af
(4007c0 is the ld1b in the sha512-sve)
Reviewed-by: Richard Henderson <richard.henderson@linaro.org>
Cc: Robert Henry <robhenry@microsoft.com>
Cc: Aaron Lindsay <aaron@os.amperecomputing.com>
Signed-off-by: Alex Bennée <alex.bennee@linaro.org>
Message-Id: <20230630180423.558337-20-alex.bennee@linaro.org>
Add an x-rme cpu property to enable FEAT_RME.
Add an x-l0gptsz property to set GPCCR_EL3.L0GPTSZ,
for testing various possible configurations.
We're not currently completely sure whether FEAT_RME will
be OK to enable purely as a CPU-level property, or if it will
need board co-operation, so we're making these experimental
x- properties, so that the people developing the system
level software for RME can try to start using this and let
us know how it goes. The command line syntax for enabling
this will change in future, without backwards-compatibility.
Signed-off-by: Richard Henderson <richard.henderson@linaro.org>
Message-id: 20230620124418.805717-21-richard.henderson@linaro.org
Reviewed-by: Peter Maydell <peter.maydell@linaro.org>
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
We use the user_ss[] array to hold the user emulation sources,
and the softmmu_ss[] array to hold the system emulation ones.
Hold the latter in the 'system_ss[]' array for parity with user
emulation.
Mechanical change doing:
$ sed -i -e s/softmmu_ss/system_ss/g $(git grep -l softmmu_ss)
Signed-off-by: Philippe Mathieu-Daudé <philmd@linaro.org>
Reviewed-by: Richard Henderson <richard.henderson@linaro.org>
Message-Id: <20230613133347.82210-10-philmd@linaro.org>
Signed-off-by: Richard Henderson <richard.henderson@linaro.org>
Convert the instructions in the load/store exclusive (STXR,
STLXR, LDXR, LDAXR) and load/store ordered (STLR, STLLR,
LDAR, LDLAR) to decodetree.
Note that for STLR, STLLR, LDAR, LDLAR this fixes an under-decoding
in the legacy decoder where we were not checking that the RES1 bits
in the Rs and Rt2 fields were set.
The new function ldst_iss_sf() is equivalent to the existing
disas_ldst_compute_iss_sf(), but it takes the pre-decoded 'ext' field
rather than taking an undecoded two-bit opc field and extracting
'ext' from it. Once all the loads and stores have been converted
to decodetree disas_ldst_compute_iss_sf() will be unused and
can be deleted.
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
Reviewed-by: Richard Henderson <richard.henderson@linaro.org>
Message-id: 20230602155223.2040685-9-peter.maydell@linaro.org
Convert the exception generation instructions SVC, HVC, SMC, BRK and
HLT to decodetree.
The old decoder decoded the halting-debug insnns DCPS1, DCPS2 and
DCPS3 just in order to then make them UNDEF; as with DRPS, we don't
bother to decode them, but document the patterns in a64.decode.
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
Reviewed-by: Richard Henderson <richard.henderson@linaro.org>
Message-id: 20230602155223.2040685-8-peter.maydell@linaro.org
In the recent refactoring we missed a few places which should be
calling finalize_memop_asimd() for ASIMD loads and stores but
instead are just calling finalize_memop(); fix these.
For the disas_ldst_single_struct() and disas_ldst_multiple_struct()
cases, this is not a behaviour change because there the size
is never MO_128 and the two finalize functions do the same thing.
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
Reviewed-by: Richard Henderson <richard.henderson@linaro.org>
In disas_ldst_reg_imm9() we missed one place where a call to
a gen_mte_check* function should now be passed the memop we
have created rather than just being passed the size. Fix this.
Fixes: 0a9091424d ("target/arm: Pass memop to gen_mte_check1*")
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
Reviewed-by: Richard Henderson <richard.henderson@linaro.org>
Reviewed-by: Philippe Mathieu-Daudé <philmd@linaro.org>
The LDG instruction loads the tag from a memory address (identified
by [Xn + offset]), and then merges that tag into the destination
register Xt. We implemented this correctly for the case when
allocation tags are enabled, but didn't get it right when ATA=0:
instead of merging the tag bits into Xt, we merged them into the
memory address [Xn + offset] and then set Xt to that.
Merge the tag bits into the old Xt value, as they should be.
Cc: qemu-stable@nongnu.org
Fixes: c15294c1e3 ("target/arm: Implement LDG, STG, ST2G instructions")
Reviewed-by: Richard Henderson <richard.henderson@linaro.org>
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
The atomic memory operations are supposed to return the old memory
data value in the destination register. This value is not
sign-extended, even if the operation is the signed minimum or
maximum. (In the pseudocode for the instructions the returned data
value is passed to ZeroExtend() to create the value in the register.)
We got this wrong because we were doing a 32-to-64 zero extend on the
result for 8 and 16 bit data values, rather than the correct amount
of zero extension.
Fix the bug by using ext8u and ext16u for the MO_8 and MO_16 data
sizes rather than ext32u.
Cc: qemu-stable@nongnu.org
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
Reviewed-by: Richard Henderson <richard.henderson@linaro.org>
Message-id: 20230602155223.2040685-2-peter.maydell@linaro.org
