Deletion of global properties puts 'the hole' in the global property
cell and updates the property details in the property dictionary with
the information that the property has been deleted. When setting
global properties that have been deleted in generated code we just
store the new value in the global property cell. This does not update
the property details in the property dictionary. Therefore, it looks
like the property is not there eventhough it was just reintroduced.
Perform 'the hole' checks in generated code for global property stores
and bail out of ICs and optimized code if storing to a property cell
that contains 'the hole'.
Review URL: http://codereview.chromium.org/6306014
git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@6508 ce2b1a6d-e550-0410-aec6-3dcde31c8c00
To avoid deopts a few extra changes were needed:
o Enable megamorphic state for special property loads on
primitives. We used to flip between monomorphic stubs.
o Extract pure string (no string wrapper support) version of the
string length stub.
Review URL: http://codereview.chromium.org/6334015
git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@6472 ce2b1a6d-e550-0410-aec6-3dcde31c8c00
conversions for external array types, which implement the Typed Array
spec. (Revision of http://codereview.chromium.org/6315004 .)
Prefer SSE2 code path on x86 processors. Non-SSE2 processors now make
a slow runtime call for float-to-int conversions. Use SSE3 for 32-bit
signed and unsigned int array types where possible.
The movement of code from ic-arm.cc to stub-cache-arm.cc caused the
VFP3 code path to be tested for the first time. Fixed bugs in the
register usage and in the constant value stored into integer arrays
for NaN and +/-Infinity.
Added new truncation test to test-api.cc. Storage of NaN and +/-Inf
was already covered. Ran unit tests on x86, x64 and ARM simulator.
Tested ia32 and x64 code in Chromium on Mac and Linux respectively
with Typed Array unit tests and WebGL content.
BUG=http://code.google.com/p/chromium/issues/detail?id=50972
TEST=test-api/ExternalArrays
Review URL: http://codereview.chromium.org/6303012
git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@6431 ce2b1a6d-e550-0410-aec6-3dcde31c8c00
Extra IC state is only two bits and only supported for call IC-s for
now. To change its extra state an IC stub jumps to a new miss stub
that goes to runtime as usual but then instead of going megamorphic
generates a monomorphic stub with the updated state.
Review URL: http://codereview.chromium.org/6344005
git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@6370 ce2b1a6d-e550-0410-aec6-3dcde31c8c00
inline allocation code used the expected number of properties to
calculate the number of inobject properties for an object instead of
getting the actual number from the initial map.
It is safer to use the inobject property count from the initial map in
any case because that is the amount the instances will get. I think
this disconnect got introduced when adding shrinking of objects.
Unfortuntely I haven't been able to create a simple reproduction for a
test case but this fixes the webpage that exhibits the crash. I'll see
if I can create a reproduction tomorrow.
Review URL: http://codereview.chromium.org/5278003
git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@5879 ce2b1a6d-e550-0410-aec6-3dcde31c8c00
Contextual load requires only a map check followed by a cell hole
check so we can generate pretty compact code for that. The fact that
we have inlined code is marked by mov ecx, offset instruction after
the IC call. Inlining is only enabled inside loops and in non-builtin
functions.
The generated code size increase is about 3%. This descreased the
pc-to-code cache hit rate in some of the benchmarks that trigger
GC. To compensate we now have 4 times as much entries in the cache.
Review URL: http://codereview.chromium.org/3402014
git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@5497 ce2b1a6d-e550-0410-aec6-3dcde31c8c00
Object model changes
----------------------------------------
New fixed_cow_array_map is used for the elements array of a JSObject
to mark it as COW. The JSObject's map and other fields are not
affected. The JSObject's map still has the "fast elements" bit set. It
means we can do only the receiver map check in keyed loads and the
receiver and the elements map checks in keyed stores. So introducing
COW arrays doesn't hurt performance of these operations. But note that
the elements map check is necessary in all mutating operations because
the "has fast elements" bit now means "has fast elements for reading".
EnsureWritableFastElements can be used in runtime functions to perform
the necessary lazy copying.
Generated code changes
----------------------------------------
Generic keyed load is updated to only do the receiver map check (this
could have been done earlier). FastCloneShallowArrayStub now has two
modes: clone elements and use COW elements. AssertFastElements macro
is added to check the elements when necessary. The custom call IC
generators for Array.prototype.{push,pop} are updated to avoid going
to the slow case (and patching the IC) when calling the builtin should
work.
COW enablement
----------------------------------------
Currently we only put shallow and simple literal arrays in the COW
mode. This is done by the parser.
Review URL: http://codereview.chromium.org/3144002
git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@5275 ce2b1a6d-e550-0410-aec6-3dcde31c8c00
This should help in cases like:
function Constructor() {
this.foo = constFunction;
this.bar = "baz";
}
for (...) {
o = new Constructor();
// Constant call IC will work.
o.foo();
// Inlined property load will see the same map.
use(o.bar);
}
This change also fixes a latent bug in custom call IC-s for strings
exposed by string-charcodeat.js.
Review URL: http://codereview.chromium.org/3160006
git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@5254 ce2b1a6d-e550-0410-aec6-3dcde31c8c00
These string methods can be composed from two basic blocks: charCodeAt
and fromCharCode, both of which have fast cases for certain types of
inputs. In this patch these two blocks are refactored to allow
generating the fast cases without having to jump around the slow
cases. In the slow cases since they can now be invoked both from
inline runtime functions and from IC stubs we either have to
save/restore state of the current frame or enter/leave a new internal
frame. This is handled by new RuntimeCallHelper interface. Its
implementation for virtual frame is based on FrameRegisterState class
extracted from DeferredCode class.
Review URL: http://codereview.chromium.org/2087009
git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@4733 ce2b1a6d-e550-0410-aec6-3dcde31c8c00
1) do not push receiver early---that simplifies tail call preparation
on ia32/x64 and renders special cleanup unnecessary;
2) do not do second map check if interceptor's and cached holder
are the same;
3) do not push/pop receiver if receiver and holder registers are the same
(means that receiver is interceptor's holder);
4) do batch pushes on arm;
5) minor cosmetic improvements.
Review URL: http://codereview.chromium.org/2282001
git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@4732 ce2b1a6d-e550-0410-aec6-3dcde31c8c00
- New сardmarking write barrier handles large objects and normal objects in a similar fashion (no more additional space for pointer tracking is required, no conditional branches in WB code).
- Changes to enable oldspaces iteration without maps decoding:
-- layout change for FixedArrays: length is stored as a smis (initial patch by
Kevin Millikin)
-- layout change for SharedFunctionInfo: integer fields are stored as smi on
arm, ia32 and rearranged on x64.
-- layout change for String: meaning of LSB bit is fliped (1 now means hash not
computed); on x64 padding is added.
-- layout of maps is _not_ changed. Map space is currently iterated in a special
way.
Review URL: http://codereview.chromium.org/2144006
git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@4715 ce2b1a6d-e550-0410-aec6-3dcde31c8c00
-- layout change for FixedArrays: length is stored as a smis (initial patch by Kevin Millikin)
-- layout change for SharedFunctionInfo: integer fields are stored as smi on arm, ia32 and rearranged on x64.
-- layout change for String: meaning of LSB bit is fliped (1 now means hash not computed); on x64 padding is added.
-- layout of maps is _not_ changed. Map space is currently iterated in a special way.
- Cardmarking write barrier. New barrier handles large objects and normal objects in a similar fashion (no more additional space for pointer tracking is required, no conditional branches in WB code).
Review URL: http://codereview.chromium.org/2101002
git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@4685 ce2b1a6d-e550-0410-aec6-3dcde31c8c00
Chromium build.
v8.gyp no longer sets any V8_TARGET_ARCH_* macro on the Mac. Instead, the
proper V8_TARGET_ARCH_* macro will be set by src/globals.h in the same way as
the V8_HOST_ARCH_* macro when it detects that no target macro is currently
defined. The Mac build will attempt to compile all ia32 and x86_64 .cc files.
#ifdef guards in each of these target-specific source files prevent their
compilation when the associated target is not selected. For completeness,
these #ifdef guards are also provided for the arm and mips .cc files.
BUG=706
TEST=x86_64 Mac GYP/Xcode-based Chromium build (still depends on other changes)
Review URL: http://codereview.chromium.org/2133003
git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@4666 ce2b1a6d-e550-0410-aec6-3dcde31c8c00
We need to be careful to check global property cells for the property
encountered during lookup. Therefore, the ICs have to be specific to
the name of the property if global objects are involved. In
principle, this means that we could get a large number of monomorphic
ICs for the same map if there is a global object in the prototype
chain. However, since this is only done for normal load ICs and not
for keyed load ICs I do not expect this to be a problem. I will
experiment with it once this goes in.
BUG=675
Review URL: http://codereview.chromium.org/1559033
git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@4426 ce2b1a6d-e550-0410-aec6-3dcde31c8c00
registration of external references in Proxy objects).
I moved the declaration of the two functions to stub-cache.h
because with all the types they use it's hard to declare them
anywhere else. But the actual definition is still in runtime.cc
near to the place where they are used.
Review URL: http://codereview.chromium.org/1079012
git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@4231 ce2b1a6d-e550-0410-aec6-3dcde31c8c00
Replaced IsValid by IsPropertyOrTransition and used IsProperty in most
of the places where IsValid was used before. Most of the time when
inspecting a lookup result we really want to know if there is a real
property present. Only for stores are we interested in transitions.
BUG=http://crbug.com/20104
TEST=cctest/test-api/NamedInterceptorMapTransitionRead
Review URL: http://codereview.chromium.org/647015
git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@3901 ce2b1a6d-e550-0410-aec6-3dcde31c8c00
When a function is called with a value type as the receiver this is now boxed as an object.
This is a low-impact solution where the receiver is only boxed when required. For IC calls to the V8 builtins values are not boxed and as most of the functions on String.prototype, Number.prototype and Boolean.prototype are sitting there most IC calls on values will not need any boxing of the receiver.
For calls which are not IC calls but calls through the CallFunctionStub a flag is used to determine whether the receiver might be a value and only when that is the case will the receiver be boxed.
No changtes to Function.call and Function.apply - they already boxed values. According to the ES5 spec the receiver should not be boxed for these functions, but current browsers have not adopted that change yet.
BUG=223
TEST=test/mjsunit/value-wrapper.js
TEST=test/mjsunit/regress/regress-crbug-3184.js
Review URL: http://codereview.chromium.org/542087
git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@3617 ce2b1a6d-e550-0410-aec6-3dcde31c8c00
callback load (or keyed load) IC.
The problem was that the IC code calls a stub, which can allocate and
thus trigger a GC if the stub is not already generated. Problem is
solved by adding the ability to "try" to call a stub, trying to
generate the stub code if necessary but signaling an allocation
failure if generating the code is not possible.
Review URL: http://codereview.chromium.org/472002
git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@3440 ce2b1a6d-e550-0410-aec6-3dcde31c8c00
The different length string types was used to encode the string length and the hash in one field. This is now split into two fields one for length and one for hash. The hash field still encodes the array index of the string if it has one. If an array index is encoded in the hash field the string length is added to the top bits of the hash field to avoid a hash value of zero.
On 32-bit this causes an additional 4 bytes to be used for all string objects. On 64-bit this will be half on average dur to pointer alignment.
Review URL: http://codereview.chromium.org/436001
git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@3350 ce2b1a6d-e550-0410-aec6-3dcde31c8c00
Change name of shifts picking the shift count from cl to sal_cl, shl_cl and shr_cl.
Add special encoding of shift by one for shr which was missing it.
git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@3314 ce2b1a6d-e550-0410-aec6-3dcde31c8c00
of individual changes:
- Added infrastructure for custom stub caching.
- Push the code object onto the stack in exit calls instead of a
debug/non-debug marker.
- Remove the DEBUG_EXIT frame type.
- Add a new exit stub generator for API getters.
Review URL: http://codereview.chromium.org/330017
git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@3130 ce2b1a6d-e550-0410-aec6-3dcde31c8c00
Objects which require an additional fixed array to be allocated now have this allocated in generated code as well. Added allocation flags to the macro assembler new space allocation routines.
Changed the ia32 and x64 macro assemblers to take allocation flags to the allocation routines instead of boolean flag.
Review URL: http://codereview.chromium.org/201015
git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@2832 ce2b1a6d-e550-0410-aec6-3dcde31c8c00
Factored out the allocation in new space from assembler code into the macro assembler. To support the current allocation patterns a number of different functions where required.
This is change is identical to http://codereview.chromium.org/174524 (committed in r2768 and reverted in r2769) except that line 710 in macro-assembler-ia32.cc is
cmp(result_end, Operand::StaticVariable(new_space_allocation_limit));
instead of
cmp(result, Operand::StaticVariable(new_space_allocation_limit));
and pre-submit fixes as well.
TBR=ager@chromium.org
Review URL: http://codereview.chromium.org/174600
git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@2771 ce2b1a6d-e550-0410-aec6-3dcde31c8c00
For objects which only have simple assignments of the form this.x = ...; a
specialized constructor stub is now generated. This generated code allocates the
object and fills in the initial properties directly. If this fails for some
reason code continues in the generic constructor stub which in turn might pass
control to the runtime system.
Added counter to see how many objects are constructed using a specialized stub.
The specialized stub is only implemented for ia32 architecture in this change.
For x64 and ARM the generic construct stub is used.
This is change is identical to http://codereview.chromium.org/174392 (committed in r2753 and reverted in r2754) except that a few parts have already been committed from http://codereview.chromium.org/173469 (committed in r2762).
Review URL: http://codereview.chromium.org/173470
git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@2764 ce2b1a6d-e550-0410-aec6-3dcde31c8c00
For objects which only have simple assignments of the form this.x = ...; a specialized constructor stub is now generated. This generated code allocates the object and fills in the initial properties directly. If this fails for some reason code continues in the generic constructor stub which in turn might pass control to the runtime system.
Added counter to see how many objects are constructed using a specialized stub.
The specialized stub is only implemented for ia32 architecture in this change. For x64 and ARM the generic construct stub is used.
Review URL: http://codereview.chromium.org/174392
git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@2753 ce2b1a6d-e550-0410-aec6-3dcde31c8c00
Remove the check for deleted properties in the global load inline cache if the property is known to be read only.
Propegate the in loop flag for the global call inline cache.
Changed the propagation of the code flags in the call stub compiler to compute these the same way for all types of call stubs and assert that the flags for the generated code is the same as those used for the cache lookup.
Addressed a few comments from previous review in test-api.cc.
Review URL: http://codereview.chromium.org/150101
git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@2308 ce2b1a6d-e550-0410-aec6-3dcde31c8c00
submitted in revisions 2093, 2094, 2099, and 2106.
There's no evidence that supports that these changes
should be the cause of the unexplained performance
regressions on the intl2 and DHTML page cyclers.
git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@2109 ce2b1a6d-e550-0410-aec6-3dcde31c8c00
called from within a loop or not. In the past we lost the
information if a call site went megamorphic before a lazily
compiled callee was called for the first time. Now we track
that correctly (this is an issue that affects richards).
We still don't manage to track the in-loop state through a
constructor call, since constructor calls use LoadICs instead
of CallICs. This issue affects delta-blue. So in this patch
we assume that lazy compilations that don't happen through a
CallIC happen from inside a loop. I have an idea to fix this
but this patch is big enough already.
With our improved tracking of in-loop state I have switched
off the inlining of in-object loads for code that is not in
a loop. This benefits compile speed. One issue is that
eagerly compiled code now doesn't get the in-object loads
inlined. We need to eagerly compile less code to fix this.
Review URL: http://codereview.chromium.org/115744
git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@2046 ce2b1a6d-e550-0410-aec6-3dcde31c8c00
This issue was raised by Brett Wilson while reviewing my changelist for readability. Craig Silverstein (one of C++ SG maintainers) confirmed that we should declare one namespace per line. Our way of namespaces closing seems not violating style guides (there is no clear agreement on it), so I left it intact.
Review URL: http://codereview.chromium.org/115756
git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@2038 ce2b1a6d-e550-0410-aec6-3dcde31c8c00
