Found by inspection, since the effect of the bug was simply to
send all memory ops through the slow path.
Signed-off-by: Richard Henderson <rth@twiddle.net>
Signed-off-by: Aurelien Jarno <aurelien@aurel32.net>
Move the slow path out of line, as the TODO's mention.
This allows the fast path to be unconditional, which can
speed up the fast path as well, depending on the core.
Signed-off-by: Richard Henderson <rth@twiddle.net>
Work better with branch predition when we have movw+movt,
as the size of the code is the same. Perhaps re-evaluate
when we have a proper constant pool.
Reviewed-by: Aurelien Jarno <aurelien@aurel32.net>
Signed-off-by: Richard Henderson <rth@twiddle.net>
The schedule was fully serial, with no possibility for dual issue.
The old schedule had a minimal issue of 7 cycles; the new schedule
has a minimal issue of 5 cycles.
Signed-off-by: Richard Henderson <rth@twiddle.net>
Share code between qemu_ld and qemu_st to process the tlb.
Reviewed-by: Aurelien Jarno <aurelien@aurel32.net>
Signed-off-by: Richard Henderson <rth@twiddle.net>
Use even more primitive helper functions to avoid lots of duplicated code.
Reviewed-by: Aurelien Jarno <aurelien@aurel32.net>
Signed-off-by: Richard Henderson <rth@twiddle.net>
Make the code more readable by only having one copy of the magic
numbers, swapping registers as needed prior to that. Speed the
compiler by not applying the rd == rn avoidance for v6 or later.
Reviewed-by: Aurelien Jarno <aurelien@aurel32.net>
Signed-off-by: Richard Henderson <rth@twiddle.net>
R12 is call clobbered, while R8 is call saved. This change
gives tcg one more call saved register for real data.
Reviewed-by: Aurelien Jarno <aurelien@aurel32.net>
Signed-off-by: Richard Henderson <rth@twiddle.net>
We get to re-use the _rIN and _rIK subroutines to handle the various
combinations of add vs sub. Fold the << 21 into the opcode enum values
so that we can explicitly add TO_CPSR as desired.
Reviewed-by: Aurelien Jarno <aurelien@aurel32.net>
Signed-off-by: Richard Henderson <rth@twiddle.net>
This greatly improves code generation for addition of small
negative constants.
Reviewed-by: Aurelien Jarno <aurelien@aurel32.net>
Signed-off-by: Richard Henderson <rth@twiddle.net>
This greatly improves the code we can produce for deposit
without armv7 support.
Reviewed-by: Aurelien Jarno <aurelien@aurel32.net>
Signed-off-by: Richard Henderson <rth@twiddle.net>
This makes it easier to verify changes to the code
generating the prologue.
[Aurelien: change the format from %i to %zu]
Reviewed-by: Aurelien Jarno <aurelien@aurel32.net>
Signed-off-by: Richard Henderson <rth@twiddle.net>
We were not allocating TCG_STATIC_CALL_ARGS_SIZE, so this meant that
any helper with more than 4 arguments would clobber the saved regs.
Realizing that we're supposed to have this memory pre-allocated means
we can clean up the tcg_out_arg functions, which were trying to do
more stack allocation.
Allocate stack memory for the TCG temporaries while we're at it.
Signed-off-by: Richard Henderson <rth@twiddle.net>
On 32-bit TCG targets, when emulating deposit_i64 with a mov_i32 +
deposit_i32, care should be taken to not overwrite the low part of
the second argument before the deposit when it is the same the
destination.
This fixes the shld instruction in qemu-system-x86_64, which in turns
fixes booting "system rescue CD version 2.8.0" on this target.
Reported-by: Michael S. Tsirkin <mst@redhat.com>
Reviewed-by: Richard Henderson <rth@twiddle.net>
Signed-off-by: Aurelien Jarno <aurelien@aurel32.net>
The TCG optimizer does great work when inserting constants, being able
to fold the open-coded deposit expansion to just an AND or an OR. Avoid
a bit the regression caused by having the deposit opcode by expanding
deposit of zero as an AND.
Reviewed-by: Aurelien Jarno <aurelien@aurel32.net>
Signed-off-by: Richard Henderson <rth@twiddle.net>
Glibc 2.16 includes an easy way to get feature bits previously
buried in /proc or the program startup auxiliary vector. Use it.
Reviewed-by: Aurelien Jarno <aurelien@aurel32.net>
Signed-off-by: Richard Henderson <rth@twiddle.net>
There are a few simple special cases that should be handled first.
Break these out to subroutines to avoid code duplication.
Reviewed-by: Aurelien Jarno <aurelien@aurel32.net>
Signed-off-by: Richard Henderson <rth@twiddle.net>
It takes half the cycles to read one CR register instead of all 8.
This is a backward compatible addition to the ISA, so chips prior
to Power 2.00 spec will simply continue to read the entire CR register.
Reviewed-by: Aurelien Jarno <aurelien@aurel32.net>
Signed-off-by: Richard Henderson <rth@twiddle.net>
Nothing else in the call chain ensures that these
constants don't have garbage in the high bits.
Reviewed-by: Aurelien Jarno <aurelien@aurel32.net>
Signed-off-by: Richard Henderson <rth@twiddle.net>
The optimization/bug being fixed is that tcg_out_cmp was not applying the
right type to loading a constant, in the case it can't be implemented
directly. Rather than recomputing the TCGType enum from the arch64 bool,
pass around the original TCGType throughout.
Reviewed-by: Aurelien Jarno <aurelien@aurel32.net>
Signed-off-by: Richard Henderson <rth@twiddle.net>
The mul_i32 pattern was loading non-16-bit constants into a register,
when we can get the middle-end to do that for us. The mul_i64 pattern
was not considering that MULLI takes 64-bit inputs.
Reviewed-by: Aurelien Jarno <aurelien@aurel32.net>
Signed-off-by: Richard Henderson <rth@twiddle.net>
Since we have special code to handle and/or/xor with a constant,
apply the same to andc/orc/eqv with a constant.
Reviewed-by: Aurelien Jarno <aurelien@aurel32.net>
Signed-off-by: Richard Henderson <rth@twiddle.net>
Using a table to look up insns of the right width and sign.
Include support for the Power 2.06 LDBRX and STDBRX insns.
Reviewed-by: Aurelien Jarno <aurelien@aurel32.net>
Signed-off-by: Richard Henderson <rth@twiddle.net>
Handle constants in common code; we'll want to reuse that later.
Reviewed-by: Aurelien Jarno <aurelien@aurel32.net>
Signed-off-by: Richard Henderson <rth@twiddle.net>
Improve constant addition -- previously we'd emit useless addi with 0.
Use new constraints to force the driver to pull full 64-bit constants
into a register.
Reviewed-by: Aurelien Jarno <aurelien@aurel32.net>
Signed-off-by: Richard Henderson <rth@twiddle.net>
We'll need a zero, and Z makes more sense for that. Make sure we
have a full compliment of signed and unsigned 16 and 32-bit tests.
Reviewed-by: Aurelien Jarno <aurelien@aurel32.net>
Signed-off-by: Richard Henderson <rth@twiddle.net>
The test for using movi32 was sub-optimal for TCG_TYPE_I32, comparing
a signed 32-bit quantity against an unsigned 32-bit quantity.
When possible, use addi+oris for 32-bit unsigned constants. Otherwise,
standardize on addi+oris+ori instead of addis+ori+rldicl.
Reviewed-by: Aurelien Jarno <aurelien@aurel32.net>
Signed-off-by: Richard Henderson <rth@twiddle.net>
We weren't ignoring the high 32 bits during a NE comparison.
Reviewed-by: Aurelien Jarno <aurelien@aurel32.net>
Signed-off-by: Richard Henderson <rth@twiddle.net>
* 'tci' of git://qemu.weilnetz.de/qemu:
tci: Make tcg temporaries local to tcg_qemu_tb_exec
tci: Delete unused tb_ret_addr
tci: Avoid code before declarations
tci: Use a local variable for env
tci: Use 32-bit signed offsets to loads/stores
We're moving away from the temporaries stored in env. Make sure we can
differentiate between temp stores and possibly bogus stores for extra
call arguments. Move TCG_AREG0 and TCG_REG_CALL_STACK out of the way
of the parameter passing registers.
Signed-off-by: Richard Henderson <rth@twiddle.net>
Signed-off by: Stefan Weil <sw@weilnetz.de>
Since the change to tcg_exit_req, the first insn of every TB is
a load with a negative offset from env.
Signed-off-by: Richard Henderson <rth@twiddle.net>
Signed-off by: Stefan Weil <sw@weilnetz.de>
When the TCG condition codes were re-organized last year,
we failed to update all of the "old-style" tests for unsigned.
Signed-off-by: Richard Henderson <rth@twiddle.net>
This can save one insn, if the constant has any bits in 32-63 set,
but no bits in 21-31 set. It never results in more insns.
Signed-off-by: Richard Henderson <rth@twiddle.net>
Since we're always in 64-bit mode, load address performs a full
64-bit add. Use that for 3-address addition, as well as for
larger constant addends when we lack extended-immediates facility.
Signed-off-by: Richard Henderson <rth@twiddle.net>
Since we have a free temporary and can always just load the constant, we
ought to do so, rather than spending the same effort constraining the const.
Signed-off-by: Richard Henderson <rth@twiddle.net>
We only support 64-bit code generation for s390x.
Don't clutter the code with ifdefs that suggest otherwise.
Signed-off-by: Richard Henderson <rth@twiddle.net>
Set TCG_TARGET_CALL_STACK_OFFSET properly for the abi. Allocate the
standard TCG_STATIC_CALL_ARGS_SIZE. And while we're at it, allocate
space for CPU_TEMP_BUF_NLONGS.
Signed-off-by: Richard Henderson <rth@twiddle.net>
In TCG, "target" means the host architecture for which TCG generates
the code. Using "guest" rather than "target" to make the document more
consistent.
Signed-off-by: Chen Wei-Ren <chenwj@iis.sinica.edu.tw>
Reviewed-by: Peter Maydell <peter.maydell@linaro.org>
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
Fix some of the nasty TCG race conditions and crashes by implementing
cpu_exit() as setting a flag which is checked at the start of each TB.
This avoids crashes if a thread or signal handler calls cpu_exit()
while the execution thread is itself modifying the TB graph (which
may happen in system emulation mode as well as in linux-user mode
with a multithreaded guest binary).
This fixes the crashes seen in LP:668799; however there are another
class of crashes described in LP:1098729 which stem from the fact
that in linux-user with a multithreaded guest all threads will
use and modify the same global TCG date structures (including the
generated code buffer) without any kind of locking. This means that
multithreaded guest binaries are still in the "unsupported"
category.
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
Reviewed-by: Richard Henderson <rth@twiddle.net>
Signed-off-by: Blue Swirl <blauwirbel@gmail.com>
Document tcg_qemu_tb_exec(). In particular, its return value is a
combination of a pointer to the next translation block and some
extra information in the low two bits. Provide some #defines for
the values passed in these bits to improve code clarity.
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
Reviewed-by: Richard Henderson <rth@twiddle.net>
Signed-off-by: Blue Swirl <blauwirbel@gmail.com>
Fix build breakage by 803d805bcef4ea7b7d6ef0b4929263e1160d6b3c:
make tcg_out_addsub2() always available.
Signed-off-by: Blue Swirl <blauwirbel@gmail.com>
Switch the default for qemu_log logging output from "/tmp/qemu.log"
to stderr. This is an incompatible change in some sense, but logging
is mostly used for debugging purposes so it shouldn't affect production
use. The previous behaviour can be obtained by adding "-D /tmp/qemu.log"
to the command line.
This change requires us to:
* update all the documentation/help text (we take the opportunity
to smooth out minor inconsistencies between the phrasing in
linux-user/bsd-user/system help messages)
* make linux-user and bsd-user defer to qemu-log for the default
logging destination rather than overriding it themselves
* ensure that all logfile closing is done via qemu_log_close()
and that that function doesn't close stderr
as well as the obvious change to the behaviour of do_qemu_set_log()
when no logfile name has been specified.
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
Reviewed-by: Stefan Hajnoczi <stefanha@redhat.com>
Reviewed-by: Markus Armbruster <armbru@redhat.com>
Message-id: 1361901160-28729-1-git-send-email-peter.maydell@linaro.org
Signed-off-by: Anthony Liguori <aliguori@us.ibm.com>
We even had the encoding of smull already handy...
Cc: Andrzej Zaborowski <balrogg@gmail.com>
Signed-off-by: Richard Henderson <rth@twiddle.net>
Signed-off-by: Blue Swirl <blauwirbel@gmail.com>
We're going to have use for this shortly in implementing other helpers.
Signed-off-by: Richard Henderson <rth@twiddle.net>
Signed-off-by: Blue Swirl <blauwirbel@gmail.com>
Commit 0b0d3320db (TCG: Final globals
clean-up) moved code_gen_prologue but forgot to update ppc code.
This broke the build on 32-bit ppc. ppc64 is unaffected.
Cc: Evgeny Voevodin <evgenyvoevodin@gmail.com>
Cc: Blue Swirl <blauwirbel@gmail.com>
Signed-off-by: Andreas Färber <andreas.faerber@web.de>
Signed-off-by: Blue Swirl <blauwirbel@gmail.com>
Rename the public-facing function cpu_set_log to qemu_set_log. This
requires us to rename the internal-only qemu_set_log() to
do_qemu_set_log().
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
Reviewed-by: Andreas Färber <afaerber@suse.de>
Signed-off-by: Blue Swirl <blauwirbel@gmail.com>
It's worth to clean-up translation blocks variables and move them
into one context as was suggested by Swirl.
Also if we use this context directly inside tcg_ctx, then it
speeds up code generation a bit.
Signed-off-by: Evgeny Voevodin <evgenyvoevodin@gmail.com>
Signed-off-by: Blue Swirl <blauwirbel@gmail.com>
Silence a (legitimate) complaint about missing parentheses:
tcg/arm/tcg-target.c: In function ‘tcg_out_qemu_ld’:
tcg/arm/tcg-target.c:1148:5: error: suggest parentheses around
comparison in operand of ‘&’ [-Werror=parentheses]
tcg/arm/tcg-target.c: In function ‘tcg_out_qemu_st’:
tcg/arm/tcg-target.c:1357:5: error: suggest parentheses around
comparison in operand of ‘&’ [-Werror=parentheses]
which meant that we would mistakenly always assert if running
a QEMU built with debug enabled on ARM.
Signed-off-by: Peter Maydell <peter.maydelL@linaro.org>
Signed-off-by: Blue Swirl <blauwirbel@gmail.com>
This adds two optimizations using the non-zero bit mask. In some cases
involving shifts or ANDs the value can become zero, and can thus be
optimized to a move of zero. Second, useless zero-extension or an
AND with constant can be detected that would only zero bits that are
already zero.
The main advantage of this optimization is that it turns zero-extensions
into moves, thus enabling much better copy propagation (around 1% code
reduction). Here is for example a "test $0xff0000,%ecx + je" before
optimization:
mov_i64 tmp0,rcx
movi_i64 tmp1,$0xff0000
discard cc_src
and_i64 cc_dst,tmp0,tmp1
movi_i32 cc_op,$0x1c
ext32u_i64 tmp0,cc_dst
movi_i64 tmp12,$0x0
brcond_i64 tmp0,tmp12,eq,$0x0
and after (without patch on the left, with on the right):
movi_i64 tmp1,$0xff0000 movi_i64 tmp1,$0xff0000
discard cc_src discard cc_src
and_i64 cc_dst,rcx,tmp1 and_i64 cc_dst,rcx,tmp1
movi_i32 cc_op,$0x1c movi_i32 cc_op,$0x1c
ext32u_i64 tmp0,cc_dst
movi_i64 tmp12,$0x0 movi_i64 tmp12,$0x0
brcond_i64 tmp0,tmp12,eq,$0x0 brcond_i64 cc_dst,tmp12,eq,$0x0
Other similar cases: "test %eax, %eax + jne" where eax is already 32-bit
(after optimization, without patch on the left, with on the right):
discard cc_src discard cc_src
mov_i64 cc_dst,rax mov_i64 cc_dst,rax
movi_i32 cc_op,$0x1c movi_i32 cc_op,$0x1c
ext32u_i64 tmp0,cc_dst
movi_i64 tmp12,$0x0 movi_i64 tmp12,$0x0
brcond_i64 tmp0,tmp12,ne,$0x0 brcond_i64 rax,tmp12,ne,$0x0
"test $0x1, %dl + je":
movi_i64 tmp1,$0x1 movi_i64 tmp1,$0x1
discard cc_src discard cc_src
and_i64 cc_dst,rdx,tmp1 and_i64 cc_dst,rdx,tmp1
movi_i32 cc_op,$0x1a movi_i32 cc_op,$0x1a
ext8u_i64 tmp0,cc_dst
movi_i64 tmp12,$0x0 movi_i64 tmp12,$0x0
brcond_i64 tmp0,tmp12,eq,$0x0 brcond_i64 cc_dst,tmp12,eq,$0x0
In some cases TCG even outsmarts GCC. :) Here the input code has
"and $0x2,%eax + movslq %eax,%rbx + test %rbx, %rbx" and the optimizer,
thanks to copy propagation, does the following:
movi_i64 tmp12,$0x2 movi_i64 tmp12,$0x2
and_i64 rax,rax,tmp12 and_i64 rax,rax,tmp12
mov_i64 cc_dst,rax mov_i64 cc_dst,rax
ext32s_i64 tmp0,rax -> nop
mov_i64 rbx,tmp0 -> mov_i64 rbx,cc_dst
and_i64 cc_dst,rbx,rbx -> nop
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
Signed-off-by: Richard Henderson <rth@twiddle.net>
Signed-off-by: Blue Swirl <blauwirbel@gmail.com>
Add a "mask" field to the tcg_temp_info struct. A bit that is zero
in "mask" will always be zero in the corresponding temporary.
Zero bits in the mask can be produced from moves of immediates,
zero-extensions, ANDs with constants, shifts; they can then be
be propagated by logical operations, shifts, sign-extensions,
negations, deposit operations, and conditional moves. Other
operations will just reset the mask to all-ones, i.e. unknown.
[rth: s/target_ulong/tcg_target_ulong/]
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
Signed-off-by: Richard Henderson <rth@twiddle.net>
Signed-off-by: Blue Swirl <blauwirbel@gmail.com>
The next patch will add to the TCG optimizer a field that should be
non-zero in the default case. Thus, replace the memset of the
temps array with a loop. Only the state field has to be up-to-date,
because others are not used except if the state is TCG_TEMP_COPY
or TCG_TEMP_CONST.
[rth: Extracted the loop to a function.]
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
Signed-off-by: Richard Henderson <rth@twiddle.net>
Signed-off-by: Blue Swirl <blauwirbel@gmail.com>
Commit 7f6f0ae5b9 added two assertions.
One of these assertions is not needed:
The pointer ts is never NULL because it is initialized with the
address of an array element.
Reviewed-by: Richard Henderson <rth@twiddle.net>
Signed-off-by: Stefan Weil <sw@weilnetz.de>
Signed-off-by: Anthony Liguori <aliguori@us.ibm.com>
Existing compile-time detection is spotty at best. Convert
it all to runtime detection instead.
Signed-off-by: Richard Henderson <rth@twiddle.net>
Signed-off-by: Blue Swirl <blauwirbel@gmail.com>
In dead_temp, local temps should always be marked as back to memory,
even if they have not been allocated (i.e. they are discared before
cross a basic block).
It fixes the following assertion in target-xtensa:
qemu-system-xtensa: tcg/tcg.c:1665: temp_save: Assertion `s->temps[temp].val_type == 2 || s->temps[temp].fixed_reg' failed.
Aborted
Reported-by: Max Filippov <jcmvbkbc@gmail.com>
Tested-by: Max Filippov <jcmvbkbc@gmail.com>
Signed-off-by: Aurelien Jarno <aurelien@aurel32.net>
The bswap16 TCG opcode assumes that the high bytes of the temp equal
to 0 before calling it. The ARM backend implementation takes this
assumption to slightly optimize the generated code.
The same implementation is called for implementing the cross-endian
qemu_st16 opcode, where this assumption is not true anymore. One way to
fix that would be to zero the high bytes before calling it. Given the
store instruction just ignore them, it is possible to provide a slightly
more optimized version. With ARMv6+ the rev16 instruction does the work
correctly. For lower ARM versions the patch provides a version which
behaves correctly with non-zero high bytes, but fill them with junk.
Cc: Andrzej Zaborowski <balrogg@gmail.com>
Cc: Peter Maydell <peter.maydell@linaro.org>
Cc: qemu-stable@nongnu.org
Reviewed-by: Peter Maydell <peter.maydell@linaro.org>
Signed-off-by: Aurelien Jarno <aurelien@aurel32.net>
The TCG arm backend considers likely that the offset to the TLB
entries does not exceed 12 bits for mem_index = 0. In practice this is
not true for at least the MIPS target.
The current patch fixes that by loading the bits 23-12 with a separate
instruction, and using loads with address writeback, independently of
the value of mem_idx. In total this allow a 24-bit offset, which is a
lot more than needed.
Cc: Andrzej Zaborowski <balrogg@gmail.com>
Cc: Peter Maydell <peter.maydell@linaro.org>
Cc: qemu-stable@nongnu.org
Signed-off-by: Aurelien Jarno <aurelien@aurel32.net>
The operations for INDEX_op_deposit_i32 and INDEX_op_deposit_i64
are now supported and enabled by default.
Signed-off-by: Stefan Weil <sw@weilnetz.de>
Signed-off-by: Blue Swirl <blauwirbel@gmail.com>
mmu access looks something like:
<check tlb>
if miss goto slow_path
<fast path>
done:
...
; end of the TB
slow_path:
<pre process>
mr r3, r27 ; move areg0 to r3
; (r3 holds the first argument for all the PPC32 ABIs)
<call mmu_helper>
b $+8
.long done
<post process>
b done
On ppc32 <call mmu_helper> is:
(SysV and Darwin)
mmu_helper is most likely not within direct branching distance from
the call site, necessitating
a. moving 32 bit offset of mmu_helper into a GPR ; 8 bytes
b. moving GPR to CTR/LR ; 4 bytes
c. (finally) branching to CTR/LR ; 4 bytes
r3 setting - 4 bytes
call - 16 bytes
dummy jump over retaddr - 4 bytes
embedded retaddr - 4 bytes
Total overhead - 28 bytes
(PowerOpen (AIX))
a. moving 32 bit offset of mmu_helper's TOC into a GPR1 ; 8 bytes
b. loading 32 bit function pointer into GPR2 ; 4 bytes
c. moving GPR2 to CTR/LR ; 4 bytes
d. loading 32 bit small area pointer into R2 ; 4 bytes
e. (finally) branching to CTR/LR ; 4 bytes
r3 setting - 4 bytes
call - 24 bytes
dummy jump over retaddr - 4 bytes
embedded retaddr - 4 bytes
Total overhead - 36 bytes
Following is done to trim the code size of slow path sections:
In tcg_target_qemu_prologue trampolines are emitted that look like this:
trampoline:
mfspr r3, LR
addi r3, 4
mtspr LR, r3 ; fixup LR to point over embedded retaddr
mr r3, r27
<jump mmu_helper> ; tail call of sorts
And slow path becomes:
slow_path:
<pre process>
<call trampoline>
.long done
<post process>
b done
call - 4 bytes (trampoline is within code gen buffer
and most likely accessible via
direct branch)
embedded retaddr - 4 bytes
Total overhead - 8 bytes
In the end the icache pressure is decreased by 20/28 bytes at the cost
of an extra jump to trampoline and adjusting LR (to skip over embedded
retaddr) once inside.
Signed-off-by: malc <av1474@comtv.ru>
Add optimized TCG qemu_ld/st generation which locates the code of TLB miss
cases at the end of a block after generating the other IRs.
Currently, this optimization supports only i386 and x86_64 hosts.
Signed-off-by: Yeongkyoon Lee <yeongkyoon.lee@samsung.com>
Signed-off-by: Blue Swirl <blauwirbel@gmail.com>
Commit 9c43b68de6 do not correctly check
for dead outputs when they need to be synced to memory in case of
half-dead operations.
Fix that by applying the same pattern than for the default case.
Tested-by: Stefan Weil <sw@weilnetz.de>
Signed-off-by: Aurelien Jarno <aurelien@aurel32.net>
MIPS32 and later instruction sets have a multiplication instruction
directly operating on GPRs. It only produces a 32-bit result but
it is exactly what is needed by QEMU.
Reviewed-by: Richard Henderson <rth@twiddle.net>
Signed-off-by: Aurelien Jarno <aurelien@aurel32.net>
When we allocate a reserved_va for the guest, the kernel will likely
choose an address well above 4G. At which point we must use a pair
of movabsq+addq to form the host address. If we have OS support,
set up a segment register to point to guest_base instead.
Signed-off-by: Richard Henderson <rth@twiddle.net>
Signed-off-by: Aurelien Jarno <aurelien@aurel32.net>
The current helper flags, TCG_CALL_CONST and TCG_CALL_PURE might be
confusing and doesn't provide enough granularity for some helpers (FP
helpers for example).
This patch changes them into the following helpers flags:
- TCG_CALL_NO_READ_GLOBALS means that the helper does not read globals,
either directly or via an exception. They will not be saved to their
canonical location before calling the helper.
- TCG_CALL_NO_WRITE_GLOBALS means that the helper does not modify any
globals. They will only be saved to their canonical locations before
calling helpers, but they won't be reloaded afterwise.
- TCG_CALL_NO_SIDE_EFFECTS means that the call to the function is
removed if the return value is not used.
It provides convenience flags, to avoid helper definitions longer than
80 characters. It also provides compatibility flags, and updates the
documentation.
Reviewed-by: Richard Henderson <rth@twiddle.net>
Signed-off-by: Aurelien Jarno <aurelien@aurel32.net>
Operations with side effects (in practice qemu_ld/st ops), only need to
synchronize globals to make sure the CPU state is consistent in case of
exception.
Reviewed-by: Richard Henderson <rth@twiddle.net>
Signed-off-by: Aurelien Jarno <aurelien@aurel32.net>
Mapping a memory address using a global and accessing it through
ld/st operations is currently broken. As it doesn't make any sense
to do that performance wise, let's forbid that.
Update the TCG documentation, and remove partial support for that.
Reviewed-by: Richard Henderson <rth@twiddle.net>
Signed-off-by: Aurelien Jarno <aurelien@aurel32.net>
Some branch related ops are marked with TCG_OPF_SIDE_EFFECTS, some other
not. In practice they don't need to, as they are all marked with
TCG_OPF_BB_END, which is handled specifically in all the code.
The call op is marked as TCG_OPF_SIDE_EFFECTS, which might be not true
as there is are specific flags (TCG_CALL_CONST and TCG_CALL_PURE) for
specifying that. On the other hand it always clobber arguments, so mark
it as such even if the call op is handled in a different code path.
Reviewed-by: Richard Henderson <rth@twiddle.net>
Signed-off-by: Aurelien Jarno <aurelien@aurel32.net>
The liveness analysis ensures that globals and temps are at the correct
state at a basic block end or with an op with side effects. Avoid
looping on all temps, this can be time consuming on targets with a lot
of globals. Keep an assert in debug mode.
Reviewed-by: Richard Henderson <rth@twiddle.net>
Signed-off-by: Aurelien Jarno <aurelien@aurel32.net>
Start with local temps in TEMP_VAL_MEM state, to make possible a later
check that all the temps are correctly saved back to memory.
Reviewed-by: Richard Henderson <rth@twiddle.net>
Signed-off-by: Aurelien Jarno <aurelien@aurel32.net>
Always mark dead input arguments as dead, even if the op is at the basic
block end. This will allow to check that all temps are correctly saved.
Reviewed-by: Richard Henderson <rth@twiddle.net>
Signed-off-by: Aurelien Jarno <aurelien@aurel32.net>
Now that the liveness analysis provides more information, rewrite
tcg_reg_alloc_mov(). This changes the behaviour about propagating
constants and memory accesses. We now take the assumption that once
a value is loaded into a register (from memory or from a constant),
it's better to keep it there than to reload it later. This assumption
is now always almost correct given that we are now sure the
corresponding temp is going to be used later (otherwise it would have
been synchronized and marked as dead already). The assumption is wrong
if one of the op after clobbers some registers including the one
of the holding the temp (this can be avoided by allocating clobbered
registers last, which is what most TCG target do), or in case of lack
of available register.
Reviewed-by: Richard Henderson <rth@twiddle.net>
Signed-off-by: Aurelien Jarno <aurelien@aurel32.net>
Now that the liveness analysis might mark some output temps as dead, call
temp_dead() if needed.
Reviewed-by: Richard Henderson <rth@twiddle.net>
Signed-off-by: Aurelien Jarno <aurelien@aurel32.net>
Rework the liveness analysis by tracking temps that need to go back to
memory in addition to dead temps tracking. This allows to mark output
arguments as "need sync", and to synchronize them back to memory as soon
as they are not written anymore. This way even arguments mapping to
globals can be marked as "dead", avoiding moves to a new register when
input and outputs are aliased.
In addition it means that registers are freed as soon as temps are not
used anymore, instead of waiting for a basic block end or an op with side
effects. This reduces register spilling especially on CPUs with few
registers, and spread the mov over all the TB, increasing the
performances on in-order CPUs.
Reviewed-by: Richard Henderson <rth@twiddle.net>
Signed-off-by: Aurelien Jarno <aurelien@aurel32.net>
Synchronize an output argument when requested by the liveness analysis.
This is needed so that the temp can be declared dead later.
For that, add a new op_sync_args table in which each bit tells if the
corresponding output argument needs to be synchronized with the memory.
Pass it to the tcg_reg_alloc_* functions, and honor this bit. We need to
synchronize the argument before marking it as dead, and we have to make
sure all the infos about the temp are correctly filled.
At the same time change some types from unsigned int to uint16_t when
passing op_dead_args.
Reviewed-by: Richard Henderson <rth@twiddle.net>
Signed-off-by: Aurelien Jarno <aurelien@aurel32.net>
Add a new function temp_sync() to synchronize the canonical location
of a temp with the value in the corresponding register, but without
freeing the associated register. Rewrite temp_save() to call
temp_sync() followed by temp_dead().
Reviewed-by: Richard Henderson <rth@twiddle.net>
Signed-off-by: Aurelien Jarno <aurelien@aurel32.net>
Add a new function tcg_reg_sync() to synchronize the canonical location
of a temp with the value in the associated register, but without freeing
it. Rewrite tcg_reg_free() to first call tcg_reg_sync() and then to free
the register.
Reviewed-by: Richard Henderson <rth@twiddle.net>
Signed-off-by: Aurelien Jarno <aurelien@aurel32.net>
A lot of code is duplicated to mark a temporary as dead. Replace it
by temp_dead(), which in addition marks the temp as saved in memory
for globals and local temps, instead of doing this a posteriori in
temp_save().
Reviewed-by: Richard Henderson <rth@twiddle.net>
Signed-off-by: Aurelien Jarno <aurelien@aurel32.net>
On x86_64, remove the constraint on the third argument register which
is not needed:
- For loads the helper arguments are env, addr, mem_idx. The addr
value should not be in the two first argument registers as they are
used in tcg_out_tlb_load().
- For stores the helper arguments are env, addr, data, mem_idx.
The addr and data values should not be in the two first argument
registers as they are used in tcg_out_tlb_load(). The data value
should also not be in the two first argument registers, but could
be in the third argument register in which case it would be already
loaded at the right location.
Reviewed-by: Richard Henderson <rth@twiddle.net>
Signed-off-by: Aurelien Jarno <aurelien@aurel32.net>
Now that CONFIG_TCG_PASS_AREG0 has been removed, it's easier to get
an optimal code for the load/store functions.
First swap the two registers used in tcg_out_tlb_load() so that the
address end-up in the second register instead of the first one. Adjust
tcg_out_qemu_ld() and tcg_out_qemu_st() to respectively call
tcg_out_qemu_ld_direct() and tcg_out_qemu_st_direct() with the correct
registers. Then replace the register shifting by direct load of the
arguments.
Reviewed-by: Richard Henderson <rth@twiddle.net>
Signed-off-by: Aurelien Jarno <aurelien@aurel32.net>
We had a hack for arm and sparc, allocating code_gen_prologue to a
special section. Which, honestly does no good under certain cases.
We've already got limits on code_gen_buffer_size to ensure that all
TBs can use direct branches between themselves; reuse this limit to
ensure the prologue is also reachable.
As a bonus, we get to avoid marking a page of the main executable's
data segment as executable.
Signed-off-by: Richard Henderson <rth@twiddle.net>
Signed-off-by: Blue Swirl <blauwirbel@gmail.com>
Like add2, do operand ordering, constant folding, and dead operand
elimination. The latter happens about 15% of all mulu2 during an
x86_64 bios boot.
Signed-off-by: Richard Henderson <rth@twiddle.net>
Signed-off-by: Aurelien Jarno <aurelien@aurel32.net>
When x86_64 guest is not in 64-bit mode, the high-part of the 64-bit
add is dead. When the host is 32-bit, we can simplify to 32-bit
arithmetic.
Signed-off-by: Richard Henderson <rth@twiddle.net>
Signed-off-by: Aurelien Jarno <aurelien@aurel32.net>