gwj / at91sam9263 · Commits

Optimize code generation in gen_muldiv():
- Don't do sign extension when the value is already guaranteed to be
  sign extended (otherwise, results are marked as UNPREDICTABLE).
- Access the LO, HI registers directly instead of writting them through
  a temporary variable.

Signed-off-by: Aurelien Jarno <aurelien@aurel32.net>

git-svn-id: svn://svn.savannah.nongnu.org/qemu/trunk@5681 c046a42c-6fe2-441c-8c8c-71466251a162

Optimize code generation in gen_arith()/gen_arith_imm():
- Don't do sign extension when the value is already guaranteed to be
  sign extended (otherwise, results are marked as UNPREDICTABLE).
- When the value is sign extended, compare the value to 0 instead of
  testing bit 31/63.
- Temp variables are valid up to and *including* the brcond instruction.
  Use them instead of temp local variables.

Signed-off-by: Aurelien Jarno <aurelien@aurel32.net>

git-svn-id: svn://svn.savannah.nongnu.org/qemu/trunk@5680 c046a42c-6fe2-441c-8c8c-71466251a162

Bit shuffle operations can be written with very few TCG instructions
(between 5 and 8), so it is worth converting them to TCG.

This code also move all bit shuffle generation code to a separate
function in order to have a cleaner exception code path, that is it
doesn't store back the TCG register to the target register after the
exception, as the TCG register doesn't exist anymore.

Signed-off-by: Aurelien Jarno <aurelien@aurel32.net>

git-svn-id: svn://svn.savannah.nongnu.org/qemu/trunk@5679 c046a42c-6fe2-441c-8c8c-71466251a162

Bitfield operations can be written with very few TCG instructions
(between 2 and 5), so it is worth converting them to TCG.

Signed-off-by: Aurelien Jarno <aurelien@aurel32.net>

git-svn-id: svn://svn.savannah.nongnu.org/qemu/trunk@5678 c046a42c-6fe2-441c-8c8c-71466251a162

Instead of dynamically generating different code depending on the UX
flag, add a new flag in ctx->flags to generate different code.

Signed-off-by: Aurelien Jarno <aurelien@aurel32.net>

git-svn-id: svn://svn.savannah.nongnu.org/qemu/trunk@5677 c046a42c-6fe2-441c-8c8c-71466251a162

The user mode can be tested at translation time using ctx->hflags.
This simplifies gen_op_addr_add().

Signed-off-by: Aurelien Jarno <aurelien@aurel32.net>

git-svn-id: svn://svn.savannah.nongnu.org/qemu/trunk@5676 c046a42c-6fe2-441c-8c8c-71466251a162

We obviously don't need to use a temporary variable to write PC.

Signed-off-by: Aurelien Jarno <aurelien@aurel32.net>

git-svn-id: svn://svn.savannah.nongnu.org/qemu/trunk@5675 c046a42c-6fe2-441c-8c8c-71466251a162

This patch attempts to fix mft* helpers and the associated TCG calls.
mft* helpers do not take a register in argument, however:
- some helpers are called with an argument while they do not take one.
- some helpers are declared with an argument they don't use.

Acked-by: Thiemo Seufer <ths@networkno.de>
Signed-off-by: Aurelien Jarno <aurelien@aurel32.net>

git-svn-id: svn://svn.savannah.nongnu.org/qemu/trunk@5674 c046a42c-6fe2-441c-8c8c-71466251a162

Move tcg_temp_free() out of the conditional part to make sure
the TCG temporary variable is freed in all cases.

Signed-off-by: Aurelien Jarno <aurelien@aurel32.net>
Acked-by: Thiemo Seufer <ths@networkno.de>

git-svn-id: svn://svn.savannah.nongnu.org/qemu/trunk@5673 c046a42c-6fe2-441c-8c8c-71466251a162

On Alpha the target to native fcntl definitions were missing.  Because of
this, programs trying to open files with the O_CREAT option were getting
O_APPEND instead, etc.

This was keeping gcc from the spec benchmarks from running, among other
things.

(Vince Weaver)

git-svn-id: svn://svn.savannah.nongnu.org/qemu/trunk@5672 c046a42c-6fe2-441c-8c8c-71466251a162

Noticed by Chris Krumme.

Signed-off-by: Aurelien Jarno <aurelien@aurel32.net>

git-svn-id: svn://svn.savannah.nongnu.org/qemu/trunk@5671 c046a42c-6fe2-441c-8c8c-71466251a162

LD/STD/LWA require displacement to be multiple of 4, provide
tcg_out_ldsta which checks the supplied displacement and falls
back on indexed variant when the check fails. All uses of
LD/STD/LWA outside of tcg_out_ldst appear to be safe.

git-svn-id: svn://svn.savannah.nongnu.org/qemu/trunk@5670 c046a42c-6fe2-441c-8c8c-71466251a162

We don't need to use cpu_loop_exit() because we never use the 
condition codes so everything can be folded into a single case.

Signed-off-by: Anthony Liguori <aliguori@us.ibm.com>



git-svn-id: svn://svn.savannah.nongnu.org/qemu/trunk@5669 c046a42c-6fe2-441c-8c8c-71466251a162

Signed-off-by: Aurelien Jarno <aurelien@aurel32.net>

git-svn-id: svn://svn.savannah.nongnu.org/qemu/trunk@5668 c046a42c-6fe2-441c-8c8c-71466251a162

The cmpbge instruction should compare all 8 bytes of one 64-bit value with
another.  However, we were looping with a < 7 condition which was skipping
the top byte.  So if we were doing a compare where the top byte was
important, we could get the wrong result (this notably breaks the strlen()
function with certain sized strings).

(Vince Weaver)

git-svn-id: svn://svn.savannah.nongnu.org/qemu/trunk@5667 c046a42c-6fe2-441c-8c8c-71466251a162

env->*dt.base should fit target address space, so we should use
target_mmap to allocate them.

Signed-off-by: Kirill A. Shutemov <kirill@shutemov.name>


git-svn-id: svn://svn.savannah.nongnu.org/qemu/trunk@5666 c046a42c-6fe2-441c-8c8c-71466251a162

result in a cryptic error (and an abort):
  qemu: block-vvfat.c:97: array_get: Assertion `index < array->next' failed.
  Aborted

Turn this into an actually useful error message:
  Directory does not fit in FAT16 (capacity 504MB)
  qemu: could not open disk image fat:$DIR/


git-svn-id: svn://svn.savannah.nongnu.org/qemu/trunk@5665 c046a42c-6fe2-441c-8c8c-71466251a162

x86 CPUs feature extended family/model bits in CPUID leaf
0000_0001|EAX. Refer to page 10 in:
http://www.amd.com/us-en/assets/content_type/white_papers_and_tech_docs/25481.pdf

Those bits are necessary to model newer AMD CPUs:
-cpu qemu64,family=15,model=65,stepping=3 or
-cpu qemu64,family=16,model=4,stepping=2

Signed-off-by: Andre Przywara <andre.przywara@amd.com>


git-svn-id: svn://svn.savannah.nongnu.org/qemu/trunk@5664 c046a42c-6fe2-441c-8c8c-71466251a162

git-svn-id: svn://svn.savannah.nongnu.org/qemu/trunk@5663 c046a42c-6fe2-441c-8c8c-71466251a162

Also fix which argument gets negated in fandnot[12] and fornot[12]


git-svn-id: svn://svn.savannah.nongnu.org/qemu/trunk@5662 c046a42c-6fe2-441c-8c8c-71466251a162

Signed-off-by: Aurelien Jarno <aurelien@aurel32.net>

git-svn-id: svn://svn.savannah.nongnu.org/qemu/trunk@5661 c046a42c-6fe2-441c-8c8c-71466251a162

Signed-off-by: Aurelien Jarno <aurelien@aurel32.net>

git-svn-id: svn://svn.savannah.nongnu.org/qemu/trunk@5660 c046a42c-6fe2-441c-8c8c-71466251a162

- Rename to gen_op_arith_neg for consistency with other functions.
- Correctly free TCG temp variable.
- Fix the return value in 64-bit mode in case of overflow.

Signed-off-by: Aurelien Jarno <aurelien@aurel32.net>

git-svn-id: svn://svn.savannah.nongnu.org/qemu/trunk@5659 c046a42c-6fe2-441c-8c8c-71466251a162

gen_op_arith_divw():
- "deoptimize" gen_op_arith_divw to make it more readable.
- Correctly free TCG temp variable

gen_op_arith_divd():
- Call the right function.

Signed-off-by: Aurelien Jarno <aurelien@aurel32.net>

git-svn-id: svn://svn.savannah.nongnu.org/qemu/trunk@5658 c046a42c-6fe2-441c-8c8c-71466251a162

Signed-off-by: Aurelien Jarno <aurelien@aurel32.net>

git-svn-id: svn://svn.savannah.nongnu.org/qemu/trunk@5657 c046a42c-6fe2-441c-8c8c-71466251a162

Signed-off-by: Aurelien Jarno <aurelien@aurel32.net>

git-svn-id: svn://svn.savannah.nongnu.org/qemu/trunk@5656 c046a42c-6fe2-441c-8c8c-71466251a162

git-svn-id: svn://svn.savannah.nongnu.org/qemu/trunk@5655 c046a42c-6fe2-441c-8c8c-71466251a162

git-svn-id: svn://svn.savannah.nongnu.org/qemu/trunk@5654 c046a42c-6fe2-441c-8c8c-71466251a162

Also optimise qemu_strdup by using memcpy - using pstrcpy is usually 
suboptimal.


git-svn-id: svn://svn.savannah.nongnu.org/qemu/trunk@5653 c046a42c-6fe2-441c-8c8c-71466251a162

git-svn-id: svn://svn.savannah.nongnu.org/qemu/trunk@5652 c046a42c-6fe2-441c-8c8c-71466251a162

The use of strncat and strndup was correct, pstrcpy and pstrdup wasn't.
I'll try to restore building on non-gnu OSes in a later commit.


git-svn-id: svn://svn.savannah.nongnu.org/qemu/trunk@5651 c046a42c-6fe2-441c-8c8c-71466251a162

Generate an option rom instead of using a hijacked boot sector for kernel
booting.  This just requires adding a small option ROM header and a few more
instructions to the boot sector to take over the int19 vector and run our
boot code.

A disk is no longer needed when using -kernel on x86.

Signed-off-by: Anthony Liguori <aliguori@us.ibm.com>



git-svn-id: svn://svn.savannah.nongnu.org/qemu/trunk@5650 c046a42c-6fe2-441c-8c8c-71466251a162

Signed-off-by: Aurelien Jarno <aurelien@aurel32.net>

git-svn-id: svn://svn.savannah.nongnu.org/qemu/trunk@5649 c046a42c-6fe2-441c-8c8c-71466251a162

It looks like the i386 runs out of registers for allocation due
to too many global registers allocated by the ppc target.

Here is a quick and dirty fix that seems to solve the problem.
This should be considered as temporary.

Signed-off-by: Laurent Desnogues <laurent.desnogues@gmail.com
Signed-off-by: Aurelien Jarno <aurelien@aurel32.net>

git-svn-id: svn://svn.savannah.nongnu.org/qemu/trunk@5648 c046a42c-6fe2-441c-8c8c-71466251a162

git-svn-id: svn://svn.savannah.nongnu.org/qemu/trunk@5647 c046a42c-6fe2-441c-8c8c-71466251a162

Signed-off-by: Christoph Egger <Christoph.Egger@amd.com>


git-svn-id: svn://svn.savannah.nongnu.org/qemu/trunk@5646 c046a42c-6fe2-441c-8c8c-71466251a162

Fix reading of cpu_lock in gen_qemu_stql_c, original patch from Laurent
Desnogues.

A new flag was added to gen_store_mem to allocate local temps instead
of temps;  this flag should be set when the tcg_gen_qemu_store callback
uses brcond before using the temps or else liveness analysis will get
rid of the temps.

This also adds lock printing in cpu_dump_state which can help
debug.

git-svn-id: svn://svn.savannah.nongnu.org/qemu/trunk@5645 c046a42c-6fe2-441c-8c8c-71466251a162

Signed-off-by: Aurelien Jarno <aurelien@aurel32.net>

git-svn-id: svn://svn.savannah.nongnu.org/qemu/trunk@5644 c046a42c-6fe2-441c-8c8c-71466251a162

Commit #5620 revealed an issue of the SSTEP_NOIRQ masking that was
applied on all interrupt sources (including internal ones) when single
stepping through the guest. Due to that commit, we now ended up in an
infinite loop when CPU_INTERRUPT_EXIT was pending on SSTEP resume. That
was due to #5620 eating all TBs while CPU_INTERRUPT_EXIT is pending, but
SSTEP_NOIRQ preventing CPU_INTERRUPT_EXIT to be processed.

What SSTEP_NOIRQ should actually do is to block the delivery of all
external, guest visible interrupts. With the fix below applied, single
stepping now works again.

Signed-off-by: Jan Kiszka <jan.kiszka@siemens.com>

git-svn-id: svn://svn.savannah.nongnu.org/qemu/trunk@5643 c046a42c-6fe2-441c-8c8c-71466251a162

- Makefile.target:  re-enable profiling for user qemu.  It seems
  profiling was (accidently?) removed by commit 3937

- syscall.c:
    * add an include to get _mcleanup prototype
    * add a call to _mcleanup for exit_group in a way
      similar to what is done for exit

(Laurent Desnogues)

git-svn-id: svn://svn.savannah.nongnu.org/qemu/trunk@5642 c046a42c-6fe2-441c-8c8c-71466251a162