All of these pass the current mask to FunctionEmitContext::SetBlockEntryMask()
so that when a break/continue/return is encountered, it can test to see if all
lanes have followed that path and then return; this in turn ensures that we never
run statements with an all-off execution mask.
These functions were passing the function internal mask, not the full mask, and
thus could end up executing code with the mask all off if some lanes were
disabled by an outer function. (The new tests test this case.)
When "break", "continue", or "return" is used under varying control flow,
we now always check the execution mask to see if all of the program
instances are executing it. (Previously, this was only done with "cbreak",
"ccontinue", and "creturn", which are now deprecated.)
An important effect of this change is that it fixes a family of cases
where we could end up running with an "all off" execution mask, which isn't
supposed to happen, as it leads to all sorts of invalid behavior.
This change does cause the volume rendering example to run 9% slower, but
doesn't affect the other examples.
Issue #257.
This fixes a crash if 'cbreak' was used in a 'switch'. Renamed
FunctionEmitContext::SetLoopMask() to SetBlockEntryMask(), and
similarly the loopMask member variable.
We need to do this since it's illegal to have nested foreach statements, but
nested foreach_unique, or foreach_unique inside foreach, etc., are all fine.
When the outermost dimension(s) were partially active, but the innermost
dimension was all on, we'd inadvertently use an incorrect "all on"
execution mask.
Fixes issues #177 and #200.
If we have a simple varying 'if' statement where the only code in the body is
a single 'break', then emit special case code that just updates the execution
mask directly.
Surprisingly, this leads to better generated code (e.g. Mandelbrot 7.1x on AVX
vs 5.8x before). It's not clear why the general code generation path for
break doesn't generate the equivalent code; this topic should be investigated
further. (Issue #277).
It can sometimes be useful to know the general place we were in the program
when an assertion hit; when the position is available / applicable, this
macro is now used.
Issue #268.
We now have a set of template functions CastType<AtomicType>, etc., that in
turn use a new typeId field in each Type instance, allowing them to be inlined
and to be quite efficient.
This improves front-end performance for a particular large program by 28%.
Both ReturnStmt and DeclStmt now check the values being associated
with references to make sure that they are legal (e.g. it's illegal
to assign a varying lvalue, or a compile-time constant to a reference
type). Previously we didn't catch this and would end up hitting
assertions in LLVM when code did this stuff.
Mostly fixes issue #225 (except for adding a FAQ about what this
error message means.)
We now have separate Expr implementations for dereferencing pointers
and automatically dereferencing references. This is in particular
necessary so that we can detect attempts to dereference references
with the '*' operator in programs and issue an error in that case.
Fixes issue #192.
For now this has the __ prefix, as an experimental feature currently only
used in the standard library implementation. It's probably worth making
something along these lines an official feature, but I'm not sure if this
in its current form is quite the right thing.
With AOS data, we can often coalesce the accesses into gathers for the main
part of foreach loops but only fail on the last bits where the mask is not
all on (since the coalescing code doesn't handle mixed masks, yet.) Before,
we'd report success with coalescing and then also report that gathers were needed
for the same accesses that were coalesced, which was a) confusing, and b)
didn't accurately represent what was going on for the majority of the loop
iterations.
In particular, we 1. weren't setting the function mask to 'all on', such that
any mixed function mask would in turn apply inside the foreach loop, and 2.
weren't always setting the internal mask to 'all on' before doing any additional
masking based on the iteration variables.
Previously, we uniqued AtomicTypes, so that they could be compared
by pointer equality, but with forthcoming SOA variability changes,
this would become too unwieldy (lacking a more general / ubiquitous
type uniquing implementation.)
We now follow C's approach of evaluating these: we don't evaluate
the second expression in the operator if the value of the first one
determines the overall result. Thus, these can now be used
idiomatically like (index < limit && array[index] > 0) and such.
For varying expressions, the mask is set appropriately when evaluating
the second expression.
(For expressions that can be determined to be both simple and safe to
evaluate with the mask all off, we still evaluate both sides and compute
the logical op result directly, which saves a number of branches and tests.
However, the effect of this should never be visible to the programmer.)
Issue #4.
Previously, when we had a switch statement with a uniform switch condition
but a 'break' statement that was under varying control flow inside the
switch, we'd promote the switch condition to be varying so that the
break would work correctly.
Now, we leave the condition as uniform and are thus able to use the
more-efficient LLVM switch instruction in this case.
Issue #156.
Specialize the code for the innermost loop to not do any masking
computations for the innermost dimension for the iterations where
we are certainly working on a full vector's worth of data.
This fix improves performance/code quality of "foreach" such that
it's essentially the same as the equivalent "for" loop.
Fixes issue #151.
Switches with both uniform and varying "switch" expressions are
supported. Switch statements with varying expressions and very
large numbers of labels may not perform well; some issues to be
filed shortly will track opportunities for improving these.
Previously, we would return immediately if the current basic block
was NULL; however, this is the wrong thing to do in that goto labels
and case/default labels in switch statements will establish a new
current basic block even if the current one is NULL.
Now, when a type is declared without an explicit "uniform" or "varying"
qualifier, its variability is unbound; depending on the context of the
declaration, the variability is later finalized.
Currently, in almost all cases, types with unbound variability are
resolved to varying types; the one exception is typecasts like:
"(int)1"; in this case, the fact that (int) has unbound variability
carries through to the TypeCastExpr, which in turn notices that the
expression being type cast has uniform type and in turn will resolve
(int) to (uniform int).
Fixes issue #127.
ispc now supports goto, but only under uniform control flow--i.e.
it must be possible for the compiler to statically determine that
all program instances will follow the goto. An error is issued at
compile time if a goto is used when this is not the case.
When used, these targets end up with calls to undefined functions for all
of the various special vector stuff ispc needs to compile ispc programs
(masked store, gather, min/max, sqrt, etc.).
These targets are not yet useful for anything, but are a step toward
having an option to C++ code with calls out to intrinsics.
Reorganized the directory structure a bit and put the LLVM bitcode used
to define target-specific stuff (as well as some generic built-ins stuff)
into a builtins/ directory.
Note that for building on Windows, it's now necessary to set a LLVM_VERSION
environment variable (with values like LLVM_2_9, LLVM_3_0, LLVM_3_1svn, etc.)
Specifically, stmts and exprs are no longer responsible for first recursively
optimizing their children before doing their own optimization (this turned
out to be error-prone, with children sometimes being forgotten.) They now
are just responsible for their own optimization, when appropriate.
