Provide a common routine for the places that require ALIGN(PC, 4)
as the base address as opposed to plain PC. The two are always
the same for A32, but the difference is meaningful for thumb mode.
Backports commit 16e0d8234ef9291747332d2c431e46808a060472 from qemu
We currently have 3 different ways of computing the architectural
value of "PC" as seen in the ARM ARM.
The value of s->pc has been incremented past the current insn,
but that is all. Thus for a32, PC = s->pc + 4; for t32, PC = s->pc;
for t16, PC = s->pc + 2. These differing computations make it
impossible at present to unify the various code paths.
With the newly introduced s->pc_curr, we can compute the correct
value for all cases, using the formula given in the ARM ARM.
This changes the behaviour for load_reg() and load_reg_var()
when called with reg==15 from a 32-bit Thumb instruction:
previously they would have returned the incorrect value
of pc_curr + 6, and now they will return the architecturally
correct value of PC, which is pc_curr + 4. This will not
affect well-behaved guest software, because all of the places
we call these functions from T32 code are instructions where
using r15 is UNPREDICTABLE. Using the architectural PC value
here is more consistent with the T16 and A32 behaviour.
Backports commit fdbcf6329d0c2984c55d7019419a72bf8e583c36 from qemu
Add a new field to retain the address of the instruction currently
being translated. The 32-bit uses are all within subroutines used
by a32 and t32. This will become less obvious when t16 support is
merged with a32+t32, and having a clear definition will help.
Convert aarch64 as well for consistency. Note that there is one
instance of a pre-assert fprintf that used the wrong value for the
address of the current instruction.
Backports commit 43722a6d4f0c92f7e7e1e291580039b0f9789df1 from qemu
This function is used in two different contexts, and it will be
clearer if the function is given the address to which it applies.
Backports commit 331b1ca616cb708db30dab68e3262d286e687f24 from qemu
When generating an architectural single-step exception we were
routing it to the "default exception level", which is to say
the same exception level we execute at except that EL0 exceptions
go to EL1. This is incorrect because the debug exception level
can be configured by the guest for situations such as single
stepping of EL0 and EL1 code by EL2.
We have to track the target debug exception level in the TB
flags, because it is dependent on CPU state like HCR_EL2.TGE
and MDCR_EL2.TDE. (That we were previously calling the
arm_debug_target_el() function to determine dc->ss_same_el
is itself a bug, though one that would only have manifested
as incorrect syndrome information.) Since we are out of TB
flag bits unless we want to expand into the cs_base field,
we share some bits with the M-profile only HANDLER and
STACKCHECK bits, since only A-profile has this singlestep.
Fixes: https://bugs.launchpad.net/qemu/+bug/1838913
Backports commit 8bd587c1066f4456ddfe611b571d9439a947d74c from qemu
Factor out code to 'generate a singlestep exception', which is
currently repeated in four places.
To do this we need to also pull the identical copies of the
gen-exception() function out of translate-a64.c and translate.c
into translate.h.
(There is a bug in the code: we're taking the exception to the wrong
target EL. This will be simpler to fix if there's only one place to
do it.)
Backports commit c1d5f50f094ab204accfacc2ee6aafc9601dd5c4 from qemu
An attempt to do an exception-return (branch to one of the magic
addresses) in linux-user mode for M-profile should behave like
a normal branch, because linux-user mode is always going to be
in 'handler' mode. This used to work, but we broke it when we added
support for the M-profile security extension in commit d02a8698d7ae2bfed.
In that commit we allowed even handler-mode calls to magic return
values to be checked for and dealt with by causing an
EXCP_EXCEPTION_EXIT exception to be taken, because this is
needed for the FNC_RETURN return-from-non-secure-function-call
handling. For system mode we added a check in do_v7m_exception_exit()
to make any spurious calls from Handler mode behave correctly, but
forgot that linux-user mode would also be affected.
How an attempted return-from-non-secure-function-call in linux-user
mode should be handled is not clear -- on real hardware it would
result in return to secure code (not to the Linux kernel) which
could then handle the error in any way it chose. For QEMU we take
the simple approach of treating this erroneous return the same way
it would be handled on a CPU without the security extensions --
treat it as a normal branch.
The upshot of all this is that for linux-user mode we should never
do any of the bx_excret magic, so the code change is simple.
This ought to be a weird corner case that only affects broken guest
code (because Linux user processes should never be attempting to do
exception returns or NS function returns), except that the code that
assigns addresses in RAM for the process and stack in our linux-user
code does not attempt to avoid this magic address range, so
legitimate code attempting to return to a trampoline routine on the
stack can fall into this case. This change fixes those programs,
but we should also look at restricting the range of memory we
use for M-profile linux-user guests to the area that would be
real RAM in hardware.
Backports commit 9027d3fba605d8f6093342ebe4a1da450d374630 from qemu
Thumb instructions in an IT block are set up to be conditionally
executed depending on a set of condition bits encoded into the IT
bits of the CPSR/XPSR. The architecture specifies that if the
condition bits are 0b1111 this means "always execute" (like 0b1110),
not "never execute"; we were treating it as "never execute". (See
the ConditionHolds() pseudocode in both the A-profile and M-profile
Arm ARM.)
This is a bit of an obscure corner case, because the only legal
way to get to an 0b1111 set of condbits is to do an exception
return which sets the XPSR/CPSR up that way. An IT instruction
which encodes a condition sequence that would include an 0b1111 is
UNPREDICTABLE, and for v8A the CONSTRAINED UNPREDICTABLE choices
for such an IT insn are to NOP, UNDEF, or treat 0b1111 like 0b1110.
Add a comment noting that we take the latter option.
Backports commit 5529de1e5512c05276825fa8b922147663fd6eac from qemu
Remove the now unused TCG globals cpu_F0s, cpu_F0d, cpu_F1s, cpu_F1d.
cpu_M0 is still used by the iwmmxt code, and cpu_V0 and
cpu_V1 are used by both iwmmxt and Neon.
Backports commit d9eea52c67c04c58ecceba6ffe5a93d1d02051fa from qemu
Remove some old constructns from NEON_2RM_VCVT_F16_F32 code:
* don't use CPU_F0s
* don't use tcg_gen_st_f32
Backports commit b66f6b9981004bbf120b8d17c20f92785179bdf2 from qemu
Remove some old constructs from NEON_2RM_VCVT_F16_F32 code:
* don't use cpu_F0s
* don't use tcg_gen_ld_f32
Backports commit 58f2682eee738e8890f9cfe858e0f4f68b00d45d from qemu
Stop using cpu_F0s for the Neon f32/s32 VCVT operations.
Since this is the last user of cpu_F0s in the Neon 2rm-op
loop, we can remove the handling code for it too.
Backports commit 60737ed5785b9c1c6f1c85575dfdd1e9eec91878 from qemu
Where Neon instructions are floating point operations, we
mostly use the old VFP utility functions like gen_vfp_abs()
which work on the TCG globals cpu_F0s and cpu_F1s. The
Neon for-each-element loop conditionally loads the inputs
into either a plain old TCG temporary for most operations
or into cpu_F0s for float operations, and similarly stores
back either cpu_F0s or the temporary.
Switch NEON_2RM_VABS_F away from using cpu_F0s, and
update neon_2rm_is_float_op() accordingly.
Backports commit fd8a68cdcf81d70eebf866a132e9780d4108da9c from qemu
Convert the float-to-integer VCVT instructions to decodetree.
Since these are the last unconverted instructions, we can
delete the old decoder structure entirely now.
Backports commit 3111bfc2da6ba0c8396dc97ca479942d711c6146 from qemu
Convert the VCVT (between floating-point and fixed-point) instructions
to decodetree.
Backports commit e3d6f4290c788e850c64815f0b3e331600a4bcc0 from qemu
Convert the VFP round-to-integer instructions VRINTR, VRINTZ and
VRINTX to decodetree.
These instructions were only introduced as part of the "VFP misc"
additions in v8A, so we check this. The old decoder's implementation
was incorrectly providing them even for v7A CPUs.
Backports commit e25155f55dc4abb427a88dfe58bbbc550fe7d643 from qemu
Convert the VCVTT and VCVTB instructions which convert from
f32 and f64 to f16 to decodetree.
Since we're no longer constrained to the old decoder's style
using cpu_F0s and cpu_F0d we can perform a direct 16 bit
store of the right half of the input single-precision register
rather than doing a load/modify/store sequence on the full
32 bits.
Backports commit cdfd14e86ab0b1ca29a702d13a8e4af2e902a9bf from qemu
Convert the VCVTT, VCVTB instructions that deal with conversion
from half-precision floats to f32 or 64 to decodetree.
Since we're no longer constrained to the old decoder's style
using cpu_F0s and cpu_F0d we can perform a direct 16 bit
load of the right half of the input single-precision register
rather than loading the full 32 bits and then doing a
separate shift or sign-extension.
Backports commit b623d803dda805f07aadcbf098961fde27315c19 from qemu
Convert the VFP comparison instructions to decodetree.
Note that comparison instructions should not honour the VFP
short-vector length and stride information: they are scalar-only
operations. This applies to all the 2-operand instructions except
for VMOV, VABS, VNEG and VSQRT. (In the old decoder this is
implemented via the "if (op == 15 && rn > 3) { veclen = 0; }" check.)
Backports commit 386bba2368842fc74388a3c1651c6c0c0c70adbd from qemu
Convert the VFP VABS instruction to decodetree.
Unlike the 3-op versions, we don't pass fpst to the VFPGen2OpSPFn or
VFPGen2OpDPFn because none of the operations which use this format
and support short vectors will need it.
Backports commit 90287e22c987e9840704345ed33d237cbe759dd9 from qemu
Convert the VFP fused multiply-add instructions (VFNMA, VFNMS,
VFMA, VFMS) to decodetree.
Note that in the old decode structure we were implementing
these to honour the VFP vector stride/length. These instructions
were introduced in VFPv4, and in the v7A architecture they
are UNPREDICTABLE if the vector stride or length are non-zero.
In v8A they must UNDEF if stride or length are non-zero, like
all VFP instructions; we choose to UNDEF always.
Backports commit d4893b01d23060845ee3855bc96626e16aad9ab5 from qemu
Convert the VFP VMLA instruction to decodetree.
This is the first of the VFP 3-operand data processing instructions,
so we include in this patch the code which loops over the elements
for an old-style VFP vector operation. The existing code to do this
looping uses the deprecated cpu_F0s/F0d/F1s/F1d TCG globals; since
we are going to be converting instructions one at a time anyway
we can take the opportunity to make the new loop use TCG temporaries,
which means we can do that conversion one operation at a time
rather than needing to do it all in one go.
We include an UNDEF check which was missing in the old code:
short-vector operations (with stride or length non-zero) were
deprecated in v7A and must UNDEF in v8A, so if the MVFR0 FPShVec
field does not indicate that support for short vectors is present
we UNDEF the operations that would use them. (This is a change
of behaviour for Cortex-A7, Cortex-A15 and the v8 CPUs, which
previously were all incorrectly allowing short-vector operations.)
Note that the conversion fixes a bug in the old code for the
case of VFP short-vector "mixed scalar/vector operations". These
happen where the destination register is in a vector bank but
but the second operand is in a scalar bank. For example
vmla.f64 d10, d1, d16 with length 2 stride 2
is equivalent to the pair of scalar operations
vmla.f64 d10, d1, d16
vmla.f64 d8, d3, d16
where the destination and first input register cycle through
their vector but the second input is scalar (d16). In the
old decoder the gen_vfp_F1_mul() operation uses cpu_F1{s,d}
as a temporary output for the multiply, which trashes the
second input operand. For the fully-scalar case (where we
never do a second iteration) and the fully-vector case
(where the loop loads the new second input operand) this
doesn't matter, but for the mixed scalar/vector case we
will end up using the wrong value for later loop iterations.
In the new code we use TCG temporaries and so avoid the bug.
This bug is present for all the multiply-accumulate insns
that operate on short vectors: VMLA, VMLS, VNMLA, VNMLS.
Note 2: the expression used to calculate the next register
number in the vector bank is not in fact correct; we leave
this behaviour unchanged from the old decoder and will
fix this bug later in the series.
Backports commit 266bd25c485597c94209bfdb3891c1d0c573c164 from qemu
Expand out the sequences in the new decoder VLDR/VSTR/VLDM/VSTM trans
functions which perform the memory accesses by going via the TCG
globals cpu_F0s and cpu_F0d, to use local TCG temps instead.
Backports commit 3993d0407dff7233e42f2251db971e126a0497e9 from qemu
Convert the VFP load/store multiple insns to decodetree.
This includes tightening up the UNDEF checking for pre-VFPv3
CPUs which only have D0-D15 : they now UNDEF for any access
to D16-D31, not merely when the smallest register in the
transfer list is in D16-D31.
This conversion does not try to share code between the single
precision and the double precision versions; this looks a bit
duplicative of code, but it leaves the door open for a future
refactoring which gets rid of the use of the "F0" registers
by inlining the various functions like gen_vfp_ld() and
gen_mov_F0_reg() which are hiding "if (dp) { ... } else { ... }"
conditionalisation.
Backports commit fa288de272c5c8a66d5eb683b123706a52bc7ad6 from qemu
Convert the VFP two-register transfer instructions to decodetree
(in the v8 Arm ARM these are the "Advanced SIMD and floating-point
64-bit move" encoding group).
Again, we expand out the sequences involving gen_vfp_msr() and
gen_msr_vfp().
Backports commit 81f681106eabe21c55118a5a41999fb7387fb714 from qemu
Convert the "single-precision" register moves to decodetree:
* VMSR
* VMRS
* VMOV between general purpose register and single precision
Note that the VMSR/VMRS conversions make our handling of
the "should this UNDEF?" checks consistent between the two
instructions:
* VMSR to MVFR0, MVFR1, MVFR2 now UNDEF from EL0
(previously was a nop)
* VMSR to FPSID now UNDEFs from EL0 or if VFPv3 or better
(previously was a nop)
* VMSR to FPINST and FPINST2 now UNDEF if VFPv3 or better
(previously would write to the register, which had no
guest-visible effect because we always UNDEF reads)
We also tighten up the decode: we were previously underdecoding
some SBZ or SBO bits.
The conversion of VMOV_single includes the expansion out of the
gen_mov_F0_vreg()/gen_vfp_mrs() and gen_mov_vreg_F0()/gen_vfp_msr()
sequences into the simpler direct load/store of the TCG temp via
neon_{load,store}_reg32(): we know in the new function that we're
always single-precision, we don't need to use the old-and-deprecated
cpu_F0* TCG globals, and we don't happen to have the declaration of
gen_vfp_msr() and gen_vfp_mrs() at the point in the file where the
new function is.
Backports commit a9ab50011aeda2dd012da99069e078379315ea18 from qemu
Convert the "double-precision" register moves to decodetree:
this covers VMOV scalar-to-gpreg, VMOV gpreg-to-scalar and VDUP.
Note that the conversion process has tightened up a few of the
UNDEF encoding checks: we now correctly forbid:
* VMOV-to-gpr with U:opc1:opc2 == 10x00 or x0x10
* VMOV-from-gpr with opc1:opc2 == 0x10
* VDUP with B:E == 11
* VDUP with Q == 1 and Vn<0> == 1
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
---
The accesses of elements < 32 bits could be improved by doing
direct ld/st of the right size rather than 32-bit read-and-shift
or read-modify-write, but we leave this for later cleanup,
since this series is generally trying to stick to fixing
the decode.
Backports commit 9851ed9269d214c0c6feba960dd14ff09e6c34b4 from qemu
