TGT_LE and TGT_BE are not size dependent and do not need to be
redefined. The others are no longer used at all.
Reviewed-by: Emilio G. Cota <cota@braap.org>
Reviewed-by: Alex Bennée <alex.bennee@linaro.org>
Signed-off-by: Richard Henderson <rth@twiddle.net>
Saves 2k code size off of a cold path.
Reviewed-by: Emilio G. Cota <cota@braap.org>
Reviewed-by: Alex Bennée <alex.bennee@linaro.org>
Signed-off-by: Richard Henderson <rth@twiddle.net>
We already include exec/address-spaces.h and exec/memory.h in
cputlb.c; the include of qemu/timer.h appears to be a fossil.
Reviewed-by: Emilio G. Cota <cota@braap.org>
Reviewed-by: Alex Bennée <alex.bennee@linaro.org>
Signed-off-by: Richard Henderson <rth@twiddle.net>
The return address argument to the softmmu template helpers was
confused. In the legacy case, we wanted to indicate that there
is no return address, and so passed in NULL. However, we then
immediately subtracted GETPC_ADJ from NULL, resulting in a non-zero
value, indicating the presence of an (invalid) return address.
Push the GETPC_ADJ subtraction down to the only point it's required:
immediately before use within cpu_restore_state_from_tb, after all
NULL pointer checks have been completed.
This makes GETPC and GETRA identical. Remove GETRA as the lesser
used macro, replacing all uses with GETPC.
Signed-off-by: Richard Henderson <rth@twiddle.net>
Previously we allowed fully unaligned operations, but not operations
that are aligned but with less alignment than the operation size.
In addition, arm32, ia64, mips, and sparc had been omitted from the
previous overalignment patch, which would have led to that alignment
being enforced.
Signed-off-by: Richard Henderson <rth@twiddle.net>
As it currently stands, QEMU does not properly handle self-modifying code
when the write is unaligned and crosses a page boundary. The procedure
for handling a write to the current translation block is to write-protect
the current translation block, catch the write, split up the translation
block into the current instruction (which remains write-protected so that
the current instruction is not modified) and the remaining instructions
in the translation block, and then restore the CPU state to before the
write occurred so the write will be retried and successfully executed.
However, since unaligned writes across pages are split into one-byte
writes for simplicity, writes to the second page (which is not the
current TB) may succeed before a write to the current TB is attempted,
and since these writes are not invalidated before resuming state after
splitting the TB, these writes will be performed a second time, thus
corrupting the second page. Credit goes to Patrick Hulin for
discovering this.
In recent 64-bit versions of Windows running in emulated mode, this
results in either being very unstable (a BSOD after a couple minutes of
uptime), or being entirely unable to boot. Windows performs one or more
8-byte unaligned self-modifying writes (xors) which intersect the end
of the current TB and the beginning of the next TB, which runs into the
aforementioned issue. This commit fixes that issue by making the
unaligned write loop perform the writes in forwards order, instead of
reverse order. This way, QEMU immediately tries to write to the current
TB, and splits the TB before any write to the second page is executed.
The write then proceeds as intended. With this patch applied, I am able
to boot and use Windows 7 64-bit and Windows 10 64-bit in QEMU without
KVM.
Per Richard Henderson's input, this patch also ensures the second page
is in the TLB before executing the write loop, to ensure the second
page is mapped.
The original discussion of the issue is located at
http://lists.nongnu.org/archive/html/qemu-devel/2014-08/msg02161.html.
Signed-off-by: Samuel Damashek <samuel.damashek@invincea.com>
Message-Id: <20160706182652.16190-1-samuel.damashek@invincea.com>
Signed-off-by: Richard Henderson <rth@twiddle.net>
[rth: Split out from the original patch.]
Signed-off-by: Samuel Damashek <samuel.damashek@invincea.com>
Message-Id: <20160706182652.16190-1-samuel.damashek@invincea.com>
Signed-off-by: Richard Henderson <rth@twiddle.net>
Some architectures (e.g. ARMv8) need the address which is aligned
to a size more than the size of the memory access.
To support such check it's enough the current costless alignment
check implementation in QEMU, but we need to support
an alignment size specifying.
Signed-off-by: Sergey Sorokin <afarallax@yandex.ru>
Message-Id: <1466705806-679898-1-git-send-email-afarallax@yandex.ru>
Signed-off-by: Richard Henderson <rth@twiddle.net>
[rth: Assert in tcg_canonicalize_memop. Leave get_alignment_bits
available for, though unused by, user-mode. Retain logging difference
based on ALIGNED_ONLY.]
Pass the MemTxAttrs for the memory access to iotlb_to_region(); this
allows it to determine the correct AddressSpace to use for the lookup.
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
Acked-by: Edgar E. Iglesias <edgar.iglesias@xilinx.com>
Now that the cpu_ld/st_* function directly call helper_ret_ld/st, we can
drop the old helper_ld/st functions.
Reviewed-by: Aurelien Jarno <aurelien@aurel32.net>
Signed-off-by: Pavel Dovgalyuk <pavel.dovgaluk@ispras.ru>
Message-Id: <20150710095656.13280.7085.stgit@PASHA-ISP>
Signed-off-by: Richard Henderson <rth@twiddle.net>
This patch introduces several helpers to pass return address
which points to the TB. Correct return address allows correct
restoring of the guest PC and icount. These functions should be used when
helpers embedded into TB invoke memory operations.
Reviewed-by: Aurelien Jarno <aurelien@aurel32.net>
Signed-off-by: Pavel Dovgalyuk <pavel.dovgaluk@ispras.ru>
Message-Id: <20150710095650.13280.32255.stgit@PASHA-ISP>
Signed-off-by: Richard Henderson <rth@twiddle.net>
After commit 626cf8f (icount: set can_do_io outside TB execution,
2014-12-08), can_do_io is set to 1 if not executing code. It is
no longer necessary to make this assumption in cpu_can_do_io.
It is also possible to remove the use_icount test, simply by
never setting cpu->can_do_io to 0 unless use_icount is true.
With these changes cpu_can_do_io boils down to a read of
cpu->can_do_io.
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
Probe for whether the specified guest write access is permitted.
If it is not permitted then an exception will be taken in the same
way as if this were a real write access (and we will not return).
Otherwise the function will return, and there will be a valid
entry in the TLB for this access.
Signed-off-by: Yongbok Kim <yongbok.kim@imgtec.com>
Reviewed-by: Leon Alrae <leon.alrae@imgtec.com>
Signed-off-by: Leon Alrae <leon.alrae@imgtec.com>
These modifiers control, on a per-memory-op basis, whether
unaligned memory accesses are allowed. The default setting
reflects the target's definition of ALIGNED_ONLY.
Reviewed-by: Peter Maydell <peter.maydell@linaro.org>
Signed-off-by: Richard Henderson <rth@twiddle.net>
The extra information is not yet used but it is now available.
This requires minor changes through all of the tcg backends.
Reviewed-by: Peter Maydell <peter.maydell@linaro.org>
Signed-off-by: Richard Henderson <rth@twiddle.net>
Add a MemTxAttrs field to the IOTLB, and allow target-specific
code to set it via a new tlb_set_page_with_attrs() function;
pass the attributes through to the device when making IO accesses.
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
Reviewed-by: Paolo Bonzini <pbonzini@redhat.com>
Reviewed-by: Edgar E. Iglesias <edgar.iglesias@xilinx.com>
Reviewed-by: Alex Bennée <alex.bennee@linaro.org>
Make the CPU iotlb a structure rather than a plain hwaddr;
this will allow us to add transaction attributes to it.
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
Reviewed-by: Paolo Bonzini <pbonzini@redhat.com>
Reviewed-by: Edgar E. Iglesias <edgar.iglesias@xilinx.com>
Reviewed-by: Alex Bennée <alex.bennee@linaro.org>
Rather than retaining io_mem_read/write as simple wrappers around
the memory_region_dispatch_read/write functions, make the latter
public and change all the callers to use them, since we need to
touch all the callsites anyway to add MemTxAttrs and MemTxResult
support. Delete io_mem_read and io_mem_write entirely.
(All the callers currently pass MEMTXATTRS_UNSPECIFIED
and convert the return value back to bool or ignore it.)
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
Reviewed-by: Alex Bennée <alex.bennee@linaro.org>
After the previous patch, TLBs will be flushed on every change to
the memory mapping. This patch augments that with synchronization
of the MemoryRegionSections referred to in the iotlb array.
With this change, it is guaranteed that iotlb_to_region will access
the correct memory map, even once the TLB will be accessed outside
the BQL.
Reviewed-by: Fam Zheng <famz@redhat.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
New MIPS features depend on the access type and enum is more convenient than
using the numbers directly.
Signed-off-by: Leon Alrae <leon.alrae@imgtec.com>
Reviewed-by: Thomas Huth <thuth@linux.vnet.ibm.com>
QEMU system mode page table walks are expensive. Taken by running QEMU
qemu-system-x86_64 system mode on Intel PIN , a TLB miss and walking a
4-level page tables in guest Linux OS takes ~450 X86 instructions on
average.
QEMU system mode TLB is implemented using a directly-mapped hashtable.
This structure suffers from conflict misses. Increasing the
associativity of the TLB may not be the solution to conflict misses as
all the ways may have to be walked in serial.
A victim TLB is a TLB used to hold translations evicted from the
primary TLB upon replacement. The victim TLB lies between the main TLB
and its refill path. Victim TLB is of greater associativity (fully
associative in this patch). It takes longer to lookup the victim TLB,
but its likely better than a full page table walk. The memory
translation path is changed as follows :
Before Victim TLB:
1. Inline TLB lookup
2. Exit code cache on TLB miss.
3. Check for unaligned, IO accesses
4. TLB refill.
5. Do the memory access.
6. Return to code cache.
After Victim TLB:
1. Inline TLB lookup
2. Exit code cache on TLB miss.
3. Check for unaligned, IO accesses
4. Victim TLB lookup.
5. If victim TLB misses, TLB refill
6. Do the memory access.
7. Return to code cache
The advantage is that victim TLB can offer more associativity to a
directly mapped TLB and thus potentially fewer page table walks while
still keeping the time taken to flush within reasonable limits.
However, placing a victim TLB before the refill path increase TLB
refill path as the victim TLB is consulted before the TLB refill. The
performance results demonstrate that the pros outweigh the cons.
some performance results taken on SPECINT2006 train
datasets and kernel boot and qemu configure script on an
Intel(R) Xeon(R) CPU E5620 @ 2.40GHz Linux machine are shown in the
Google Doc link below.
https://docs.google.com/spreadsheets/d/1eiItzekZwNQOal_h-5iJmC4tMDi051m9qidi5_nwvH4/edit?usp=sharing
In summary, victim TLB improves the performance of qemu-system-x86_64 by
11% on average on SPECINT2006, kernelboot and qemu configscript and with
highest improvement of in 26% in 456.hmmer. And victim TLB does not result
in any performance degradation in any of the measured benchmarks. Furthermore,
the implemented victim TLB is architecture independent and is expected to
benefit other architectures in QEMU as well.
Although there are measurement fluctuations, the performance
improvement is very significant and by no means in the range of
noises.
Signed-off-by: Xin Tong <trent.tong@gmail.com>
Message-id: 1407202523-23553-1-git-send-email-trent.tong@gmail.com
Reviewed-by: Peter Maydell <peter.maydell@linaro.org>
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
They do not need to be in op_helper.c. Because cputlb.c now includes
softmmu_template.h twice for each size, io_readX must be elided the
second time through.
Reviewed-by: Richard Henderson <rth@twiddle.net>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
We will reference it from more files in the next patch. To avoid
ruining the small steps we're making towards multi-target, make
it a method of CPU rather than just a global.
Reviewed-by: Andreas Färber <afaerber@suse.de>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
Do not rely on int8_t (and friends) not being preprocessor
symbols (or symbols expanding to themselves). On NetBSD (for example) the
glue(u, SDATA_TYPE) results in u__int8_t, which is undefined. There is no way
to stop cpp expanding inner macros, so just add the few lines explicitly and
get rid of the magic.
Signed-off-by: Martin Husemann <martin@NetBSD.org>
Reviewed-by: Peter Maydell <peter.maydell@linaro.org>
Reviewed-by: Andreas Färber <afaerber@suse.de>
Signed-off-by: Michael Tokarev <mjt@tls.msk.ru>
Step three in the transition: helpers not tied to the target
"default" endianness. To be used when the guest uses a memory
operation with non-default endianness.
Signed-off-by: Richard Henderson <rth@twiddle.net>
In a following patch, there will be confusion between multiple "unsigned"
suffixes; rename this one so as to imply "load".
Reviewed-by: Aurelien Jarno <aurelien@aurel32.net>
Signed-off-by: Richard Henderson <rth@twiddle.net>
Always define GETRA; use __builtin_extract_return_addr, rather than
having a special case for s390. Split GETPC_ADJ out of GETPC; use 2
universally, rather than having a special case for arm.
Rename GETPC_LDST to GETRA_LDST to indicate that it does not
contain the GETPC_ADJ value. Likewise with GETPC_EXT to GETRA_EXT.
Perform the GETPC_ADJ adjustment inside helper_ret_ld/st. This will
allow backends to pass along the "true" return address rather than
the massaged GETPC value. In the meantime, double application of
GETPC_ADJ does not hurt, since the call insn in all ISAs is at least
4 bytes long.
Reviewed-by: Aurelien Jarno <aurelien@aurel32.net>
Signed-off-by: Richard Henderson <rth@twiddle.net>
Avoid a loop in the tlb_fill path; the fill will either succeed or
generate an exception.
Inline the slow_ld/st function; it was a complete copy of the main
helper except for the actual cross-page unaligned code, and the
compiler was inlining it anyway.
Add unlikely markers optimizing for the most common case of simple
tlb miss.
Make sure the compiler can optimize away the unaligned paths for a
1 byte access.
Reviewed-by: Aurelien Jarno <aurelien@aurel32.net>
Signed-off-by: Richard Henderson <rth@twiddle.net>
Allow the code that tcg generates to be less obtuse, passing in
the return address directly instead of computing it in the helper.
Maintain the old entrance point unchanged as an alternate entry point.
Delete the helper_st*_cmmu prototypes; the implementations did not exist.
Signed-off-by: Richard Henderson <rth@twiddle.net>
The memory API is able to split it in two 4-byte accesses.
Reviewed-by: Richard Henderson <rth@twiddle.net>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
There is no reason to avoid a recompile before accessing unassigned
memory. In the end it will be treated as MMIO anyway.
Reviewed-by: Richard Henderson <rth@twiddle.net>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
It is never used, the IOTLB always goes through io_mem_notdirty.
In fact in softmmu_template.h, if it were, QEMU would crash just
below the tests, as soon as io_mem_read/write dispatches to
error_mem_read/write.
Reviewed-by: Richard Henderson <rth@twiddle.net>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
Add GETPC_EXT which is used by MMU helpers to selectively calculate the code
address of accessing guest memory when called from a qemu_ld/st optimized code
or a C function. Currently, it supports only i386 and x86-64 hosts.
Signed-off-by: Yeongkyoon Lee <yeongkyoon.lee@samsung.com>
Signed-off-by: Blue Swirl <blauwirbel@gmail.com>
target_phys_addr_t is unwieldly, violates the C standard (_t suffixes are
reserved) and its purpose doesn't match the name (most target_phys_addr_t
addresses are not target specific). Replace it with a finger-friendly,
standards conformant hwaddr.
Outstanding patchsets can be fixed up with the command
git rebase -i --exec 'find -name "*.[ch]"
| xargs s/target_phys_addr_t/hwaddr/g' origin
Signed-off-by: Avi Kivity <avi@redhat.com>
Signed-off-by: Anthony Liguori <aliguori@us.ibm.com>
Now that CONFIG_TCG_PASS_AREG0 is enabled for all targets,
remove dead code and support for !CONFIG_TCG_PASS_AREG0 case.
Remove dyngen-exec.h and all references to it. Although included by
hw/spapr_hcall.c, it does not seem to use it.
Remove unused HELPER_CFLAGS.
Signed-off-by: Blue Swirl <blauwirbel@gmail.com>
Reviewed-by: Richard Henderson <rth@twiddle.net>
Richard Henderson
Samuel Damashek
Sergey Sorokin
Peter Maydell
Pavel Dovgalyuk
Paolo Bonzini
Yongbok Kim
Leon Alrae
Xin Tong
Andreas Färber
Edgar E. Iglesias
Martin Husemann
Yeongkyoon Lee
Avi Kivity
Blue Swirl