GitLab

When compiling a function we can determine how much stack space it will
use. We therefore need to perform only a single stack check at the beginning
of a function to see if we have enough stack space. Instead of referring
directly to Sp - as we used to do in the past - the code generator uses
(old + 0) in the stack check. Stack layout phase turns (old + 0) into Sp.

The idea here is that, while we need to perform only one stack check for
each function, we could in theory place more stack checks later in the
function. They would be redundant, but not incorrect (in a sense that they
should not change program behaviour). We need to make sure however that a
stack check inserted after incrementing the stack pointer checks for a
respectively smaller stack space. This would not be the case if the code
generator produced direct references to Sp. By referencing (old + 0) we make
sure that we always check for a correct amount of stack: when converting
(old + 0) to Sp the stack layout phase takes into account changes already
made to stack pointer. The idea for this change came from observations made
while debugging #8275.

Signed-off-by: Edward Z. Yang <ezyang@mit.edu>

This commit removes module StgCmmGran which has only no-op functions.
According to comments in the module, it was used by GpH, but GpH
project seems to be dead for a couple of years now.

A major cleanup of trailing whitespaces and tabs in codeGen/
directory. I also adjusted code formatting in some places.

This includes selector, ap, and constructor thunks. They are still
guarded by the -ticky-dyn-thk flag.

(This is 024df664 with a small bug fix.)

This reverts commit 024df664.

Of course I gaff on my last day...

This includes selector, ap, and constructor thunks. They are still
guarded by the -ticky-dyn-thk flag.

  * the new StgCmmArgRep module breaks a dependency cycle; I also
    untabified it, but made no real changes

  * updated the documentation in the wiki and change the user guide to
    point there

  * moved the allocation enters for ticky and CCS to after the heap check

    * I left LDV where it was, which was before the heap check at least
      once, since I have no idea what it is

  * standardized all (active?) ticky alloc totals to bytes

  * in order to avoid double counting StgCmmLayout.adjustHpBackwards
    no longer bumps ALLOC_HEAP_ctr

  * I resurrected the SLOW_CALL counters

    * the new module StgCmmArgRep breaks cyclic dependency between
      Layout and Ticky (which the SLOW_CALL counters cause)

    * renamed them SLOW_CALL_fast_<pattern> and VERY_SLOW_CALL

  * added ALLOC_RTS_ctr and _tot ticky counters

    * eg allocation by Storage.c:allocate or a BUILD_PAP in stg_ap_*_info

    * resurrected ticky counters for ALLOC_THK, ALLOC_PAP, and
      ALLOC_PRIM

    * added -ticky and -DTICKY_TICKY in ways.mk for debug ways

  * added a ticky counter for total LNE entries

  * new flags for ticky: -ticky-allocd -ticky-dyn-thunk -ticky-LNE

    * all off by default

    * -ticky-allocd: tracks allocation *of* closure in addition to
       allocation *by* that closure

    * -ticky-dyn-thunk tracks dynamic thunks as if they were functions

    * -ticky-LNE tracks LNEs as if they were functions

  * updated the ticky report format, including making the argument
    categories (more?) accurate again

  * the printed name for things in the report include the unique of
    their ticky parent as well as if they are not top-level

Top-level indirections are often generated when there is a cast, e.g.

foo :: T
foo = bar `cast` (some coercion)

For these we were generating a full-blown CAF, which is a fair chunk
of code.

This patch makes these indirections generate a single IND_STATIC
closure (4 words) instead.  This is exactly what the CAF would
evaluate to eventually anyway, we're just shortcutting the whole
process.

The Slow calling convention passes the closure in R1, but we were
ignoring this and hoping it would work, which it often did.  However,
this bug seems to have been the cause of #7192, because the
graph-colouring allocator is more sensitive to having correct liveness
information on jumps.

All Cmm procedures now include the set of global registers that are live on
procedure entry, i.e., the global registers used to pass arguments to the
procedure. Only global registers that are use to pass arguments are included in
this list.

Mostly d -> g (matching DynFlag -> GeneralFlag).
Also renamed if* to when*, matching the Haskell if/when names

The main change here is that the Cmm parser now allows high-level cmm
code with argument-passing and function calls.  For example:

foo ( gcptr a, bits32 b )
{
  if (b > 0) {
     // we can make tail calls passing arguments:
     jump stg_ap_0_fast(a);
  }

  return (x,y);
}

More details on the new cmm syntax are in Note [Syntax of .cmm files]
in CmmParse.y.

The old syntax is still more-or-less supported for those occasional
code fragments that really need to explicitly manipulate the stack.
However there are a couple of differences: it is now obligatory to
give a list of live GlobalRegs on every jump, e.g.

  jump %ENTRY_CODE(Sp(0)) [R1];

Again, more details in Note [Syntax of .cmm files].

I have rewritten most of the .cmm files in the RTS into the new
syntax, except for AutoApply.cmm which is generated by the genapply
program: this file could be generated in the new syntax instead and
would probably be better off for it, but I ran out of enthusiasm.

Some other changes in this batch:

 - The PrimOp calling convention is gone, primops now use the ordinary
   NativeNodeCall convention.  This means that primops and "foreign
   import prim" code must be written in high-level cmm, but they can
   now take more than 10 arguments.

 - CmmSink now does constant-folding (should fix #7219)

 - .cmm files now go through the cmmPipeline, and as a result we
   generate better code in many cases.  All the object files generated
   for the RTS .cmm files are now smaller.  Performance should be
   better too, but I haven't measured it yet.

 - RET_DYN frames are removed from the RTS, lots of code goes away

 - we now have some more canned GC points to cover unboxed-tuples with
   2-4 pointers, which will reduce code size a little.

The current fix is relatively dumb as far as where to add HpLim
checks: it will always perform a check unless we know that we're
returning from a closure or we are doing a non let-no-escape case
analysis.  The performance impact on the nofib suite looks like this:

            Min          +5.7%     -0.0%     -6.5%     -6.4%    -50.0%
            Max          +6.3%     +5.8%     +5.0%     +5.5%     +0.8%
 Geometric Mean          +6.2%     +0.1%     +0.5%     +0.5%     -0.8%

Overall, the executable bloat is the biggest problem, so we keep the old
omit-yields optimization on by default. Remember that if you need an
interruptibility guarantee, you need to recompile all of your libraries
with -fno-omit-yields.

A better fix would involve only inserting the yields necessary to break
loops; this is left as future work.
Signed-off-by: Edward Z. Yang <ezyang@mit.edu>

I've switched to passing DynFlags rather than Platform, as (a) it's
simpler to not have to extract targetPlatform in so many places, and
(b) it may be useful to have DynFlags around in future.

We weren't passing the arguments correctly to the GC functions, which
usually happened to work because the arguments were in the right
registers already.

After this fix the profiling tests go through with the new code
generator.

This saves compile time and can make a big difference in some
pathological cases (T4801)

To explicitly choose whether you want an unregisterised build you now
need to use the "--enable-unregisterised"/"--disable-unregisterised"
configure flags.

Instead of relying on common-block-elimination to share return
continuations in the common case (case-alternative heap checks) we do
it explicitly.  This isn't hard to do, is more robust, and saves some
compilation time.  Full commentary in Note [sharing continuations].

All the flags that 'ways' imply are now dynamic

By using Haskell's debugIsOn rather than CPP's "#ifdef DEBUG", we
don't need to kludge things to keep the warning checker happy etc.

The code we were generating for heap-checks in algebraic case
alternatives wasn't working well with the common-block eliminator.  A
small tweak to make the heap-check failure jump back to the same place
in all branches lets the common-block eliminator squash more code.

We were using the SRT information generated by the computeSRTs pass to
decide whether to add a static link field to a constructor or not, and
this broke when I disabled computeSRTs for the new code generator.  So
I've hacked it for now to only rely on the SRT information generated
by CoreToStg.

Also:
 - improvements to code generation: push slow-call continuations
   on the stack instead of generating explicit continuations

 - remove unused CmmInfo wrapper type (replace with CmmInfoTable)

 - squash Area and AreaId together, remove now-unused RegSlot

 - comment out old unused stack-allocation code that no longer
   compiles after removal of RegSlot

User visible changes
====================

Profilng
--------

Flags renamed (the old ones are still accepted for now):

  OLD            NEW
  ---------      ------------
  -auto-all      -fprof-auto
  -auto          -fprof-exported
  -caf-all       -fprof-cafs

New flags:

  -fprof-auto              Annotates all bindings (not just top-level
                           ones) with SCCs

  -fprof-top               Annotates just top-level bindings with SCCs

  -fprof-exported          Annotates just exported bindings with SCCs

  -fprof-no-count-entries  Do not maintain entry counts when profiling
                           (can make profiled code go faster; useful with
                           heap profiling where entry counts are not used)

Cost-centre stacks have a new semantics, which should in most cases
result in more useful and intuitive profiles.  If you find this not to
be the case, please let me know.  This is the area where I have been
experimenting most, and the current solution is probably not the
final version, however it does address all the outstanding bugs and
seems to be better than GHC 7.2.

Stack traces
------------

+RTS -xc now gives more information.  If the exception originates from
a CAF (as is common, because GHC tends to lift exceptions out to the
top-level), then the RTS walks up the stack and reports the stack in
the enclosing update frame(s).

Result: +RTS -xc is much more useful now - but you still have to
compile for profiling to get it.  I've played around a little with
adding 'head []' to GHC itself, and +RTS -xc does pinpoint the problem
quite accurately.

I plan to add more facilities for stack tracing (e.g. in GHCi) in the
future.

Coverage (HPC)
--------------

 * derived instances are now coloured yellow if they weren't used
 * likewise record field names
 * entry counts are more accurate (hpc --fun-entry-count)
 * tab width is now correct (markup was previously off in source with
   tabs)

Internal changes
================

In Core, the Note constructor has been replaced by

        Tick (Tickish b) (Expr b)

which is used to represent all the kinds of source annotation we
support: profiling SCCs, HPC ticks, and GHCi breakpoints.

Depending on the properties of the Tickish, different transformations
apply to Tick.  See CoreUtils.mkTick for details.

Tickets
=======

This commit closes the following tickets, test cases to follow:

  - Close #2552: not a bug, but the behaviour is now more intuitive
    (test is T2552)

  - Close #680 (test is T680)

  - Close #1531 (test is result001)

  - Close #949 (test is T949)

  - Close #2466: test case has bitrotted (doesn't compile against current
    version of vector-space package)

And some knock-on changes