RPCS3/llvm-mirror
1
0
Fork 0
mirror of https://github.com/RPCS3/llvm-mirror.git synced 2024-10-28 22:42:52 +01:00
Code Issues Projects Releases Wiki Activity
36,142 Commits 4 Branches 5 Tags 704 MiB
db4f51e1b9
Commit Graph

2243 Commits

Author SHA1 Message Date
Gordon Henriksen
db4f51e1b9 With this patch, the LowerGC transformation becomes the 
ShadowStackCollector, which additionally has reduced overhead with
no sacrifice in portability.

Considering a function @fun with 8 loop-local roots,
ShadowStackCollector introduces the following overhead
(x86):

; shadowstack prologue
        movl    L_llvm_gc_root_chain$non_lazy_ptr, %eax
        movl    (%eax), %ecx
        movl    $___gc_fun, 20(%esp)
        movl    $0, 24(%esp)
        movl    $0, 28(%esp)
        movl    $0, 32(%esp)
        movl    $0, 36(%esp)
        movl    $0, 40(%esp)
        movl    $0, 44(%esp)
        movl    $0, 48(%esp)
        movl    $0, 52(%esp)
        movl    %ecx, 16(%esp)
        leal    16(%esp), %ecx
        movl    %ecx, (%eax)

; shadowstack loop overhead
        (none)

; shadowstack epilogue
        movl    48(%esp), %edx
        movl    %edx, (%ecx)

; shadowstack metadata
        .align  3
___gc_fun:                              # __gc_fun
        .long   8
        .space  4

In comparison to LowerGC:

; lowergc prologue
        movl    L_llvm_gc_root_chain$non_lazy_ptr, %eax
        movl    (%eax), %ecx
        movl    %ecx, 48(%esp)
        movl    $8, 52(%esp)
        movl    $0, 60(%esp)
        movl    $0, 56(%esp)
        movl    $0, 68(%esp)
        movl    $0, 64(%esp)
        movl    $0, 76(%esp)
        movl    $0, 72(%esp)
        movl    $0, 84(%esp)
        movl    $0, 80(%esp)
        movl    $0, 92(%esp)
        movl    $0, 88(%esp)
        movl    $0, 100(%esp)
        movl    $0, 96(%esp)
        movl    $0, 108(%esp)
        movl    $0, 104(%esp)
        movl    $0, 116(%esp)
        movl    $0, 112(%esp)

; lowergc loop overhead
        leal    44(%esp), %eax
        movl    %eax, 56(%esp)
        leal    40(%esp), %eax
        movl    %eax, 64(%esp)
        leal    36(%esp), %eax
        movl    %eax, 72(%esp)
        leal    32(%esp), %eax
        movl    %eax, 80(%esp)
        leal    28(%esp), %eax
        movl    %eax, 88(%esp)
        leal    24(%esp), %eax
        movl    %eax, 96(%esp)
        leal    20(%esp), %eax
        movl    %eax, 104(%esp)
        leal    16(%esp), %eax
        movl    %eax, 112(%esp)

; lowergc epilogue
        movl    48(%esp), %edx
        movl    %edx, (%ecx)

; lowergc metadata
        (none)

llvm-svn: 45670
 2008-01-07 01:30:53 +00:00
Duncan Sands
fd975e4b3d The transform that tries to turn calls to bitcast functions into 
direct calls bails out unless caller and callee have essentially
equivalent parameter attributes.  This is illogical - the callee's
attributes should be of no relevance here.  Rework the logic, which
incidentally fixes a crash when removed arguments have attributes.

llvm-svn: 45658
 2008-01-06 18:27:01 +00:00
Duncan Sands
b8489f09a2 When transforming a call to a bitcast function into 
a direct call with cast parameters and cast return
value (if any), instcombine was prepared to cast any
non-void return value into any other, whether castable
or not.  Add a new predicate for testing whether casting
is valid, and check it both for the return value and
(as a cleanup) for the parameters.

llvm-svn: 45657
 2008-01-06 10:12:28 +00:00
Chris Lattner
7e1c3aa702 remove a couple more unsafe xforms in the face of overflow. 
llvm-svn: 45613
 2008-01-05 01:22:42 +00:00
Chris Lattner
983697dfac remove the (x-y) < 0 comparison xform, it miscompiles 
things that are not equality comparisons, for example:
   (2147479553+4096)-2147479553 < 0    !=   (2147479553+4096) < 2147479553

llvm-svn: 45612
 2008-01-05 01:18:20 +00:00
Wojciech Matyjewicz
9ec15f974f fix typo 
llvm-svn: 45594
 2008-01-04 20:02:18 +00:00
Chris Lattner
ad9a6ccb83 Remove attribution from file headers, per discussion on llvmdev. 
llvm-svn: 45418
 2007-12-29 20:36:04 +00:00
Chris Lattner
8193d4af33 remove attribution from lib Makefiles. 
llvm-svn: 45415
 2007-12-29 20:09:26 +00:00
Christopher Lamb
dfad5f19b4 Disable null pointer folding transforms for non-generic address spaces. This should probably be a target-specific predicate based on address space. That way for targets where this isn't applicable the predicate can be optimized away. 
llvm-svn: 45403
 2007-12-29 07:56:53 +00:00
Owen Anderson
ebd3e9c500 Repair a transform that Chris noticed a bug in. Thanks to Nicholas for pointing out my stupid mistakes when writing this patch. :-) 
llvm-svn: 45384
 2007-12-28 07:42:12 +00:00
Chris Lattner
2456399ce5 disable this instcombine xform, it miscompiles: 
define i32 @main() {
entry:
	%z = alloca i32		; <i32*> [#uses=2]
	store i32 0, i32* %z
	%tmp = load i32* %z		; <i32> [#uses=1]
	%sub = sub i32 %tmp, 1		; <i32> [#uses=1]
	%cmp = icmp ult i32 %sub, 0		; <i1> [#uses=1]
	%retval = select i1 %cmp, i32 1, i32 0		; <i32> [#uses=1]
	ret i32 %retval
}

into ret 1, instead of ret 0.

Christopher, please investigate.

llvm-svn: 45383
 2007-12-28 06:24:31 +00:00
Chris Lattner
90df7f7424 Don't break critical edges for single-bb loops, this helps with PR1877, though 
it is only a partial fix.  This change is noise for most programs, but
speeds up Shootout-C++/matrix by 20%, Ptrdist/ks by 24%, smg2000 by 8%,
hexxagon by 9%, bzip2 by 9% (not sure I trust this), ackerman by 13%, etc.

OTOH, it slows down Shootout/fib2 by 40% (I'll update PR1877 with this info).

llvm-svn: 45354
 2007-12-25 19:06:45 +00:00
Chris Lattner
7e1e1f2933 add a -backedge-hack llc-beta option to codegenprepare. 
When specified, don't split backedges of single-bb loops.
This helps address PR1877

llvm-svn: 45344
 2007-12-24 19:32:55 +00:00
Chris Lattner
d64df490ca implement InstCombine/shift-trunc-shift.ll. This allows 
us to compile:
#include <math.h>
int t1(double d) { return signbit(d); }

into:

_t1:
	movd	%xmm0, %rax
	shrq	$63, %rax
	ret

instead of:

_t1:
	movd	%xmm0, %rax
	shrq	$32, %rax
	shrl	$31, %eax
	ret

on x86-64.

llvm-svn: 45311
 2007-12-22 09:07:47 +00:00
Christopher Lamb
7ca648a7b1 Implement review feedback, including additional transforms 
(icmp slt (sub A B) 1) -> (icmp sle A B)
icmp sgt (sub A B) -1) -> (icmp sge A B)

and add testcase.

llvm-svn: 45256
 2007-12-20 07:21:11 +00:00
Evan Cheng
eb07401701 Clean up previous patch: PHI uses should not prevent iv reuse if all other uses are addresses. This trades a constant multiply for one fewer iv. 
llvm-svn: 45251
 2007-12-20 02:20:53 +00:00
Chris Lattner
1a386cbdae simplify this code with the new m_Zero() pattern. Make sure the select only 
has a single use, and generalize it to not require N to be a constant.

llvm-svn: 45250
 2007-12-20 01:56:58 +00:00
Evan Cheng
cf7b6b419a Allow iv reuse if the user is a PHI node which is in turn used as addresses. 
llvm-svn: 45230
 2007-12-19 23:33:23 +00:00
Duncan Sands
56f3add5b7 When inlining through an 'nounwind' call, mark inlined 
calls 'nounwind'.  It is important for correct C++
exception handling that nounwind markings do not get
lost, so this transformation is actually needed for
correctness.

llvm-svn: 45218
 2007-12-19 21:13:37 +00:00
Christopher Lamb
be0cbc7e92 Fold subtracts into integer compares vs. zero. This improves generate code for this case on X86 
from
_foo:
        movl    $99, %ecx
        movl    4(%esp), %eax
        subl    %eax, %ecx
        xorl    %edx, %edx
        testl   %ecx, %ecx
        cmovs   %edx, %eax
        ret

to
_foo:
        xorl    %ecx, %ecx
        movl    4(%esp), %eax
        cmpl    $99, %eax
        cmovg   %ecx, %eax
        ret

llvm-svn: 45173
 2007-12-18 21:32:20 +00:00
Christopher Lamb
051a1320e8 Fix comments 
llvm-svn: 45170
 2007-12-18 20:33:11 +00:00
Christopher Lamb
d56318b885 Remove an orthogonal transformation of the selection condition from my most recent submission. 
llvm-svn: 45169
 2007-12-18 20:30:28 +00:00
Duncan Sands
242f80be86 Rename isNoReturn to doesNotReturn, and isNoUnwind to 
doesNotThrow.

llvm-svn: 45160
 2007-12-18 09:59:50 +00:00
Christopher Lamb
437b4d229e Fix typos. 
llvm-svn: 45159
 2007-12-18 09:45:40 +00:00
Christopher Lamb
aeb76743dc Fold certain additions through selects (and their compares) so as to eliminate subtractions. This code is often produced by the SMAX expansion in SCEV. 
This implements test/Transforms/InstCombine/2007-12-18-AddSelCmpSub.ll

llvm-svn: 45158
 2007-12-18 09:34:41 +00:00
David Greene
3f3a521a33 Get rid of annoying spaces. 
llvm-svn: 45100
 2007-12-17 17:40:29 +00:00
David Greene
8fda00ca05 Fix GLIBCXX_DEBUG errors. Erase invalidates std::vector iterators 
passed the erased element.

llvm-svn: 45099
 2007-12-17 17:39:51 +00:00
Christopher Lamb
a608afb52e Change the PointerType api for creating pointer types. The old functionality of PointerType::get() has become PointerType::getUnqual(), which returns a pointer in the generic address space. The new prototype of PointerType::get() requires both a type and an address space. 
llvm-svn: 45082
 2007-12-17 01:12:55 +00:00
Duncan Sands
bf62f62058 Make instcombine promote inline asm calls to 'nounwind' 
calls.  Remove special casing of inline asm from the
inliner.  There is a potential problem: the verifier
rejects invokes of inline asm (not sure why).  If an
asm call is not marked "nounwind" in some .ll, and
instcombine is not run, but the inliner is run, then
an illegal module will be created.  This is bad but
I'm not sure what the best approach is.  I'm tempted
to remove the check in the verifier...

llvm-svn: 45073
 2007-12-16 15:51:49 +00:00
Evan Cheng
4acb4eb95e Fix typo. 
llvm-svn: 44997
 2007-12-13 07:50:36 +00:00
Evan Cheng
a152909956 Be extra careful with extension use optimation. Now turned on by default. 
llvm-svn: 44981
 2007-12-13 03:32:53 +00:00
Wojciech Matyjewicz
8bb1d9e67c 1. "Upgrage" comments. 
2. Using zero-extended value of Scale and unsigned division is safe provided
   that Scale doesn't have the sign bit set.
   Previously these 2 instructions:
        %p = bitcast [100 x {i8,i8,i8}]* %x to i8*
        %q = getelementptr i8* %p, i32 -4
   were combined into:
        %q = getelementptr [100 x { i8, i8, i8 }]* %x, i32 0,
               i32 1431655764, i32 0
   what was incorrect.

llvm-svn: 44936
 2007-12-12 15:21:32 +00:00
Evan Cheng
bb418c4551 Don't muck with phi nodes; bug fixes. 
llvm-svn: 44905
 2007-12-12 02:53:41 +00:00
Evan Cheng
d6ecb2e58d Bug fix. Only safe to perform extension uses optimization if the source of extension is also defined in the same BB as the extension. 
llvm-svn: 44896
 2007-12-12 00:51:06 +00:00
Duncan Sands
b7ac459292 Make PruneEH update the nounwind/noreturn attributes 
on functions as it calculates them.

llvm-svn: 44802
 2007-12-10 19:09:40 +00:00
Owen Anderson
e3de18ac1d Fix several cache coherence bugs in MemDep/GVN that were found. Also add some (disabled) debugging code 
to make such problems easier to diagnose in the future, written by Duncan Sands.

llvm-svn: 44695
 2007-12-08 01:37:09 +00:00
Chris Lattner
861df2f4e9 simplify some code. 
llvm-svn: 44655
 2007-12-06 06:25:04 +00:00
Chris Lattner
1f2a96f9c1 move some ashr-specific code out of commonShiftTransforms into visitAShr. 
llvm-svn: 44650
 2007-12-06 01:59:46 +00:00
Evan Cheng
9e69c0ada8 If both result of the {s|z}xt and its source are live out, rewrite all uses of the source with result of extension. 
llvm-svn: 44643
 2007-12-05 23:58:20 +00:00
Duncan Sands
1e2e4972ff Rather than having special rules like "intrinsics cannot 
throw exceptions", just mark intrinsics with the nounwind
attribute.  Likewise, mark intrinsics as readnone/readonly
and get rid of special aliasing logic (which didn't use
anything more than this anyway).

llvm-svn: 44544
 2007-12-03 20:06:50 +00:00
Chris Lattner
0c41f32943 update file comment. 
llvm-svn: 44543
 2007-12-03 19:43:18 +00:00
Devang Patel
c47f68e747 If ExitValue operand is also defined in Loop header then 
insert new ExitValue after this operand definition.

This fixes PR1828.

llvm-svn: 44539
 2007-12-03 19:17:21 +00:00
Duncan Sands
14f11d6836 Integrate the readonly/readnone logic more deeply 
into alias analysis.  This meant updating the API
which now has versions of the getModRefBehavior,
doesNotAccessMemory and onlyReadsMemory methods
which take a callsite parameter.  These should be
used unless the callsite is not known, since in
general they can do a better job than the versions
that take a function.  Also, users should no longer
call the version of getModRefBehavior that takes
both a function and a callsite.  To reduce the
chance of misuse it is now protected.

llvm-svn: 44487
 2007-12-01 07:51:45 +00:00
Owen Anderson
3b893e1f84 Fix a miscompilation in spiff on PPC. 
llvm-svn: 44437
 2007-11-29 18:02:22 +00:00
Duncan Sands
1b0feb42e2 Add some convenience methods for querying attributes, and 
use them.

llvm-svn: 44403
 2007-11-28 17:07:01 +00:00
Duncan Sands
3602011bec Fix PR1146: parameter attributes are longer part of 
the function type, instead they belong to functions
and function calls.  This is an updated and slightly
corrected version of Reid Spencer's original patch.
The only known problem is that auto-upgrading of
bitcode files doesn't seem to work properly (see
test/Bitcode/AutoUpgradeIntrinsics.ll).  Hopefully
a bitcode guru (who might that be? :) ) will fix it.

llvm-svn: 44359
 2007-11-27 13:23:08 +00:00
Owen Anderson
43d4a82d4b Make LoopInfoBase more generic, in preparation for having MachineLoopInfo. This involves a small interface change. 
llvm-svn: 44348
 2007-11-27 03:43:35 +00:00
Owen Anderson
2aa0218a6c Fix another bug that was causing siod to fail. 
llvm-svn: 44325
 2007-11-26 07:17:19 +00:00
Owen Anderson
05275b7b14 Allow GVN to eliminate read-only function calls when it can detect that they are redundant. 
llvm-svn: 44323
 2007-11-26 02:26:36 +00:00
Anton Korobeynikov
14246a2eba Remove another leak. Due to some reason AliasSetTracker didn't had any dtor... 
llvm-svn: 44320
 2007-11-25 23:52:02 +00:00