From bugzilla-daemon at llvm.org Sun Jan 1 02:51:30 2012
From: bugzilla-daemon at llvm.org (bugzilla-daemon at llvm.org)
Date: Sun, 01 Jan 2012 08:51:30 +0000
Subject: [LLVMbugs] [Bug 11682] New: redundant overflow test not eliminated
Message-ID:
http://llvm.org/bugs/show_bug.cgi?id=11682
Bug #: 11682
Summary: redundant overflow test not eliminated
Product: libraries
Version: trunk
Platform: PC
OS/Version: Linux
Status: NEW
Severity: enhancement
Priority: P
Component: Scalar Optimizations
AssignedTo: unassignedbugs at nondot.org
ReportedBy: nicholas at mxc.ca
CC: llvmbugs at cs.uiuc.edu
Classification: Unclassified
This testcase is simplified from std::vector::_M_insert_aux, the function
called by push_back that doesn't get inlined.
declare {i8, i1} @llvm.uadd.with.overflow.i8(i8, i8)
declare void @use(i8, i1)
define void @test(i8 %a) {
%A = ashr i8 %a, 3 ; [-32..31]
%B = lshr i8 -1, 3 ; 31
%uadd = tail call { i8, i1 } @llvm.uadd.with.overflow.i8(i8 %A, i8 %A)
%sum = extractvalue { i8, i1 } %uadd, 0 ; [-64..62]
%overflow = extractvalue { i8, i1 } %uadd, 1
; %overflow is true iff %a is negative
%C = icmp ugt i8 %sum, %B
; %C is true iff %sum is [32..255]
; %C is true iff %A is [16..255]
; %C is true iff %a is [128..255]
; %C is true iff %a is negative
%D = or i1 %overflow, %C
call void @use(i8 %sum, i1 %D)
ret void
}
In this case, %C is equal to %overflow. opt -O2 does not remove the redundant
computation, which would be helpful.
--
Configure bugmail: http://llvm.org/bugs/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are on the CC list for the bug.
From bugzilla-daemon at llvm.org Sun Jan 1 06:34:08 2012
From: bugzilla-daemon at llvm.org (bugzilla-daemon at llvm.org)
Date: Sun, 01 Jan 2012 12:34:08 +0000
Subject: [LLVMbugs] [Bug 11683] New: Detecting infinite loops
Message-ID:
http://llvm.org/bugs/show_bug.cgi?id=11683
Bug #: 11683
Summary: Detecting infinite loops
Product: clang
Version: 2.9
Platform: PC
OS/Version: Linux
Status: NEW
Severity: enhancement
Priority: P
Component: Static Analyzer
AssignedTo: kremenek at apple.com
ReportedBy: angusgh at gmail.com
CC: llvmbugs at cs.uiuc.edu
Classification: Unclassified
It seems like it would be possible to cat a small subset of simple but common
infinite loops by checking whether the variables in the loop condition are
changed by either the condition check or the loop body. Sample code that this
would trigger on might look like this (yes, I've actually written this kind of
code, frequently, and it's not an error to gcc for obvious reasons).
int main()
{
int i=10;
while (i>0); //notice the semicolon
{
i--;
}
return 0;
}
This would also possibly be useful in a large set of cases where critical parts
of loops have been malformed.
--
Configure bugmail: http://llvm.org/bugs/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are on the CC list for the bug.
From bugzilla-daemon at llvm.org Sun Jan 1 08:32:22 2012
From: bugzilla-daemon at llvm.org (bugzilla-daemon at llvm.org)
Date: Sun, 01 Jan 2012 14:32:22 +0000
Subject: [LLVMbugs] [Bug 11684] New: Syntactically illegal
pointer-to-function type accepted
Message-ID:
http://llvm.org/bugs/show_bug.cgi?id=11684
Bug #: 11684
Summary: Syntactically illegal pointer-to-function type
accepted
Product: clang
Version: 3.0
Platform: All
OS/Version: All
Status: NEW
Severity: normal
Priority: P
Component: C++
AssignedTo: unassignedclangbugs at nondot.org
ReportedBy: ggreif at gmail.com
CC: dgregor at apple.com, llvmbugs at cs.uiuc.edu
Classification: Unclassified
Compile this code with clang:
------------------------------------------------------
template
struct gauge_member_cfun
{
typedef RETURN (*ref)(const RECEIVER&, ARG1) const;
template
struct regauge2
{
static RETURN ref(const RECEIVER& r, ARG1 a1) { return (r.*MEMBER_FUN)(a1);
}
};
template
ref regauge(void) const;
};
template
gauge_member_cfun gauge(RETURN (RECEIVER::*)(ARG1)
const);
struct Foo
{
char* foo1c(int) const;
};
void gaugers__HS(void)
{
gauge(&Foo::foo1c).regauge<&Foo::foo1c>();
}
------------------------------------------------------
It compiles without errors. But line 4
> typedef RETURN (*ref)(const RECEIVER&, ARG1) const;
is syntactically illegal. There can be no 'const' or 'volatile' specifier on
function pointers. So clang should reject this code, just like G++:
> gauge.cpp:4: error: ?const? and ?volatile? function specifiers on ?ref? invalid in type declaration
--
Configure bugmail: http://llvm.org/bugs/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are on the CC list for the bug.
From bugzilla-daemon at llvm.org Sun Jan 1 11:28:08 2012
From: bugzilla-daemon at llvm.org (bugzilla-daemon at llvm.org)
Date: Sun, 01 Jan 2012 17:28:08 +0000
Subject: [LLVMbugs] [Bug 11685] New: clang ignores
-funsafe-math-optimizations, -ffast-math, etc
Message-ID:
http://llvm.org/bugs/show_bug.cgi?id=11685
Bug #: 11685
Summary: clang ignores -funsafe-math-optimizations,
-ffast-math, etc
Product: clang
Version: trunk
Platform: PC
OS/Version: All
Status: NEW
Severity: normal
Priority: P
Component: Driver
AssignedTo: unassignedclangbugs at nondot.org
ReportedBy: benny.kra at gmail.com
CC: llvmbugs at cs.uiuc.edu
Classification: Unclassified
The clang driver currently doesn't pass information about unsafe fp
optimizations down into codegen (llvm-gcc got this right). Many benchmarks are
compiled with -ffast-math so this could be an easy perf win for clang.
--
Configure bugmail: http://llvm.org/bugs/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are on the CC list for the bug.
From bugzilla-daemon at llvm.org Sun Jan 1 11:43:37 2012
From: bugzilla-daemon at llvm.org (bugzilla-daemon at llvm.org)
Date: Sun, 01 Jan 2012 17:43:37 +0000
Subject: [LLVMbugs] [Bug 11686] New: Overload candidate computation is a
space hog
Message-ID:
http://llvm.org/bugs/show_bug.cgi?id=11686
Bug #: 11686
Summary: Overload candidate computation is a space hog
Product: clang
Version: trunk
Platform: PC
OS/Version: All
Status: NEW
Severity: enhancement
Priority: P
Component: C++
AssignedTo: unassignedclangbugs at nondot.org
ReportedBy: benny.kra at gmail.com
CC: dgregor at apple.com, llvmbugs at cs.uiuc.edu
Classification: Unclassified
The data structures to store conversion sequences are way too big:
StandardConversionSequence: 48 bytes
UserDefinedConversionSequence: 120 bytes
AmbiguousConversionSequence: 80 bytes
BadConversionSequence: 32 bytes
ImplicitConversionSequence: 128 bytes (union of the last 4 structs + 8 extra
bytes)
OverloadCandidate.Conversions: 544 bytes
(SmallVector)
OverloadCandidate: 752 bytes
To make matters worse we store the OverloadCandidates in a SmallVector. On a
large source file vector reallocation can easily churn through gigabytes of
memory. malloc on OSX seems not to aggressively reuse memory so we end up
wasting a ton of memory in some cases (I haven't really tracked down what's
causing this though).
Some possible solutions:
- Bit pack structs better (not many opportunities there), reduce SmallVector
size
- Pool allocate OverloadCandidates. llvm's SpecificBumpPtrAllocator doesn't
support iteration, but that can be added.
- Enhance the candidate computation to be lazier (not sure if this is feasible)
--
Configure bugmail: http://llvm.org/bugs/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are on the CC list for the bug.
From bugzilla-daemon at llvm.org Sun Jan 1 12:07:49 2012
From: bugzilla-daemon at llvm.org (bugzilla-daemon at llvm.org)
Date: Sun, 01 Jan 2012 18:07:49 +0000
Subject: [LLVMbugs] [Bug 10113] Explicit template instantiation with default
visibility fails to generate global symbols for certain types
In-Reply-To:
References:
Message-ID:
http://llvm.org/bugs/show_bug.cgi?id=10113
Rafael ?vila de Esp?ndola changed:
What |Removed |Added
----------------------------------------------------------------------------
Status|ASSIGNED |RESOLVED
Resolution| |FIXED
--- Comment #23 from Rafael ?vila de Esp?ndola 2012-01-01 12:07:49 CST ---
Fixed in r147405.
--
Configure bugmail: http://llvm.org/bugs/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are on the CC list for the bug.
From bugzilla-daemon at llvm.org Sun Jan 1 13:11:08 2012
From: bugzilla-daemon at llvm.org (bugzilla-daemon at llvm.org)
Date: Sun, 01 Jan 2012 19:11:08 +0000
Subject: [LLVMbugs] [Bug 10113] Explicit template instantiation with default
visibility fails to generate global symbols for certain types
In-Reply-To:
References:
Message-ID:
http://llvm.org/bugs/show_bug.cgi?id=10113
Ami Fischman changed:
What |Removed |Added
----------------------------------------------------------------------------
Status|RESOLVED |REOPENED
Resolution|FIXED |
--- Comment #24 from Ami Fischman 2012-01-01 13:11:08 CST ---
(In reply to comment #23)
> Fixed in r147405.
Not fixed for me. The repro from #8 still shows a difference from g++ (at
147405) and the chrome shared_library build still fails b/c of "missing"
(hidden) symbols.
--
Configure bugmail: http://llvm.org/bugs/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are on the CC list for the bug.
From bugzilla-daemon at llvm.org Sun Jan 1 13:47:44 2012
From: bugzilla-daemon at llvm.org (bugzilla-daemon at llvm.org)
Date: Sun, 01 Jan 2012 19:47:44 +0000
Subject: [LLVMbugs] [Bug 11687] New: jump-threading misses constant GEP
Message-ID:
http://llvm.org/bugs/show_bug.cgi?id=11687
Bug #: 11687
Summary: jump-threading misses constant GEP
Product: libraries
Version: trunk
Platform: PC
OS/Version: Linux
Status: NEW
Severity: enhancement
Priority: P
Component: Scalar Optimizations
AssignedTo: unassignedbugs at nondot.org
ReportedBy: nicholas at mxc.ca
CC: llvmbugs at cs.uiuc.edu
Classification: Unclassified
This is a missed optimization in std::vector _M_insert_aux, the function in
push_back() that doesn't get optimized enough to get inlined. Testcase:
declare i8** @something()
declare void @use(i8)
define void @test(i64 %a) {
entry:
%cmp = icmp eq i64 %a, 0
br i1 %cmp, label %brtrue, label %brfalse
brfalse:
%ptr = call i8** @something()
br label %brtrue
brtrue:
%phi = phi i8** [ %ptr, %brfalse ], [ null, %entry ]
%gep = getelementptr inbounds i8** %phi, i64 %a
%isnull = icmp eq i8** %gep, null
br i1 %isnull, label %brtrue2, label %brfalse2
brtrue2:
call void @use(i8 1)
ret void
brfalse2:
call void @use(i8 2)
ret void
}
The important missed optimization is the case where %a is zero. Then, %phi is
null, and %gep is "gep inbounds (null, 0)" which is then compared against null
(true) and we should thread the edge directly from %entry to %brtrue2.
This probably belongs in -jump-threading.
There may be a second missed optimization, but I'm not entirely sure. We may be
able to remove the edge from %brtrue to %brtrue2, making it branch directly to
%brfalse2. The matter at hand is whether %gep could be null, given that the GEP
is inbounds and %a is non-zero. Can we have %ptr = inttoptr(10*sizeof(i8**))
and %a = -10 without violating inbounds? I don't think so, in address space
zero. Is the only inbounds gep allowed to return null "gep inbounds (null,
all-zeros)"? Or is even that not legal? If not, then std::vector push_back() is
already being miscompiled...
--
Configure bugmail: http://llvm.org/bugs/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are on the CC list for the bug.
From bugzilla-daemon at llvm.org Sun Jan 1 13:53:12 2012
From: bugzilla-daemon at llvm.org (bugzilla-daemon at llvm.org)
Date: Sun, 01 Jan 2012 19:53:12 +0000
Subject: [LLVMbugs] [Bug 11688] New: i386-pc-win32 struct by value return
does not match MSVC calling convention
Message-ID:
http://llvm.org/bugs/show_bug.cgi?id=11688
Bug #: 11688
Summary: i386-pc-win32 struct by value return does not match
MSVC calling convention
Product: clang
Version: 3.0
Platform: PC
OS/Version: Windows NT
Status: NEW
Severity: normal
Priority: P
Component: LLVM Codegen
AssignedTo: unassignedclangbugs at nondot.org
ReportedBy: arcata at gmail.com
CC: llvmbugs at cs.uiuc.edu
Classification: Unclassified
When targeting i?86-pc-win32, Clang returns all structs by hidden pointer,
although MSVC returns structs sized {8,4,3,2,1} bytes in registers, as
documented at http://msdn.microsoft.com/en-us/library/984x0h58.aspx .
A simple test case:
---
#include
struct Struct1 {
uint64_t a;
};
struct Struct2 {
uint32_t a;
};
struct Struct3 {
char a, b, c;
};
struct Struct1 c_return_1(void) {
struct Struct1 r = { 0xC1A4C1A4C1A4C1A4ULL };
return r;
}
struct Struct2 c_return_2(void) {
struct Struct2 r = { 0xC1A4C1A4U };
return r;
}
struct Struct3 c_return_3(void) {
struct Struct3 r = { 0xAA, 0xBB, 0xCC };
return r;
}
---
cl.exe generates the following asm for c_return_1:
---
_c_return_1 PROC
; File
c:\users\joe\documents\code\clay\test\externals\abi\common\external_test1.c
; Line 15
push ebp
mov ebp, esp
sub esp, 8
; Line 16
mov DWORD PTR _r$[ebp], -1046167132 ; c1a4c1a4H
mov DWORD PTR _r$[ebp+4], -1046167132 ; c1a4c1a4H
; Line 17
mov eax, DWORD PTR _r$[ebp]
mov edx, DWORD PTR _r$[ebp+4]
; Line 18
mov esp, ebp
pop ebp
ret 0
_c_return_1 ENDP
---
You can see under "Line 17" that it moves the return value into
edx:eax. By contrast, clang expects a hidden pointer argument:
---
_c_return_1: # @c_return_1
# BB#0:
subl $12, %esp
movl 16(%esp), %eax
movsd L_c_return_1.r, %xmm0
movsd %xmm0, (%esp)
movsd %xmm0, (%eax)
addl $12, %esp
ret $4
---
Similar discrepancies occur for c_return_2 and c_return_3.
--
Configure bugmail: http://llvm.org/bugs/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are on the CC list for the bug.
From bugzilla-daemon at llvm.org Sun Jan 1 19:21:50 2012
From: bugzilla-daemon at llvm.org (bugzilla-daemon at llvm.org)
Date: Mon, 02 Jan 2012 01:21:50 +0000
Subject: [LLVMbugs] [Bug 11689] New: cmake build set my apartment on fire
Message-ID:
http://llvm.org/bugs/show_bug.cgi?id=11689
Bug #: 11689
Summary: cmake build set my apartment on fire
Product: clang
Version: unspecified
Platform: PC
OS/Version: All
Status: NEW
Severity: enhancement
Priority: P
Component: Documentation
AssignedTo: unassignedclangbugs at nondot.org
ReportedBy: nicolasweber at gmx.de
CC: llvmbugs at cs.uiuc.edu
Classification: Unclassified
1. use cmake build
actual: apartment burned down
expected: apartment stays usable.
--
Configure bugmail: http://llvm.org/bugs/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are on the CC list for the bug.
From bugzilla-daemon at llvm.org Sun Jan 1 19:23:17 2012
From: bugzilla-daemon at llvm.org (bugzilla-daemon at llvm.org)
Date: Mon, 02 Jan 2012 01:23:17 +0000
Subject: [LLVMbugs] [Bug 11689] cmake build set my apartment on fire
In-Reply-To:
References:
Message-ID:
http://llvm.org/bugs/show_bug.cgi?id=11689
Chandler Carruth changed:
What |Removed |Added
----------------------------------------------------------------------------
Status|NEW |RESOLVED
CC| |chandlerc at gmail.com
Resolution| |INVALID
--- Comment #1 from Chandler Carruth 2012-01-01 19:23:17 CST ---
Working as intended!
PS, MY MASTER PLAN IS NOW COMPLETE!!!!!
--
Configure bugmail: http://llvm.org/bugs/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are on the CC list for the bug.
From bugzilla-daemon at llvm.org Sun Jan 1 22:26:41 2012
From: bugzilla-daemon at llvm.org (bugzilla-daemon at llvm.org)
Date: Mon, 02 Jan 2012 04:26:41 +0000
Subject: [LLVMbugs] [Bug 11690] New: Support building chromium in
shared_library mode
Message-ID:
http://llvm.org/bugs/show_bug.cgi?id=11690
Bug #: 11690
Summary: Support building chromium in shared_library mode
Product: clang
Version: trunk
Platform: All
OS/Version: All
Status: NEW
Severity: enhancement
Priority: P
Component: C++
AssignedTo: unassignedclangbugs at nondot.org
ReportedBy: fischman at chromium.org
CC: dgregor at apple.com, llvmbugs at cs.uiuc.edu
Classification: Unclassified
With these env variables:
GYP_DEFINES='disable_nacl=1 enable_svg=0 proprietary_codecs=1
ffmpeg_branding=Chrome clang=1 clang_use_chrome_plugins=1
component=shared_library'
GYP_GENERATORS=ninja
GYP_GENERATOR_FLAGS='output_dir=ninja config=Debug'
build chromium with ninja on linux. The component=shared_library directive
tries to build parts of chromium as .so's and the final browser links them in
dynamically. Today this is broken because of symbol visibility problems (see
bug 10113 and bug 10038 for some history), whereas g++ does not have these
problems.
Chromium pointers:
http://code.google.com/p/chromium/wiki/LinuxBuildInstructions
http://code.google.com/p/chromium/wiki/NinjaBuild
--
Configure bugmail: http://llvm.org/bugs/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are on the CC list for the bug.
From bugzilla-daemon at llvm.org Mon Jan 2 01:50:52 2012
From: bugzilla-daemon at llvm.org (bugzilla-daemon at llvm.org)
Date: Mon, 02 Jan 2012 07:50:52 +0000
Subject: [LLVMbugs] [Bug 11677] lto lazy-loading produces invalid ir
In-Reply-To:
References:
Message-ID:
http://llvm.org/bugs/show_bug.cgi?id=11677
Rafael ?vila de Esp?ndola changed:
What |Removed |Added
----------------------------------------------------------------------------
Status|ASSIGNED |RESOLVED
Resolution| |FIXED
--- Comment #4 from Rafael ?vila de Esp?ndola 2012-01-02 01:50:52 CST ---
Fixed in r147425.
--
Configure bugmail: http://llvm.org/bugs/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are on the CC list for the bug.
From bugzilla-daemon at llvm.org Mon Jan 2 13:51:22 2012
From: bugzilla-daemon at llvm.org (bugzilla-daemon at llvm.org)
Date: Mon, 02 Jan 2012 19:51:22 +0000
Subject: [LLVMbugs] [Bug 11691] New: llc assertion failure generating arm
neon vst2 with pointer update
Message-ID:
http://llvm.org/bugs/show_bug.cgi?id=11691
Bug #: 11691
Summary: llc assertion failure generating arm neon vst2 with
pointer update
Product: new-bugs
Version: trunk
Platform: PC
OS/Version: Linux
Status: NEW
Severity: normal
Priority: P
Component: new bugs
AssignedTo: unassignedbugs at nondot.org
ReportedBy: andrew.b.adams at gmail.com
CC: llvmbugs at cs.uiuc.edu
Classification: Unclassified
Created attachment 7828
--> http://llvm.org/bugs/attachment.cgi?id=7828
bugpoint-reduced failure case
llc fails an assertion when generating the arm neon VST2 instruction when it
detects it can use the address-updating form:
llc: /data/scratch/omap4/llvm/include/llvm/CodeGen/MachineOperand.h:385:
int64_t llvm::MachineOperand::getImm() const: Assertion `isImm() && "Wrong
MachineOperand accessor"' failed.
Stack dump:
0. Program arguments: llc bugpoint-reduced-simplified.bc
1. Running pass 'Function Pass Manager' on module
'bugpoint-reduced-simplified.bc'.
2. Running pass 'If Converter' on function '@curved_inner'
Aborted
I believe this is the case because the assertion is no longer triggered if you
include this check in ARMISelLowering.cpp at line 7340
--- a/lib/Target/ARM/ARMISelLowering.cpp
+++ b/lib/Target/ARM/ARMISelLowering.cpp
@@ -7335,6 +7335,10 @@ static SDValue CombineBaseUpdate(SDNode *N,
unsigned NumVecs = 0;
if (isIntrinsic) {
unsigned IntNo = cast(N->getOperand(1))->getZExtValue();
+
+ // AA: bail out for VST2 to avoid bug
+ if (IntNo == Intrinsic::arm_neon_vst2) continue;
+
switch (IntNo) {
default: assert(0 && "unexpected intrinsic for Neon base update");
case Intrinsic::arm_neon_vld1: NewOpc = ARMISD::VLD1_UPD;
A bugpoint-simplified bc file is attached. Compile it with -march=arm
-mcpu=cortex-a9
--
Configure bugmail: http://llvm.org/bugs/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are on the CC list for the bug.
From bugzilla-daemon at llvm.org Mon Jan 2 16:02:54 2012
From: bugzilla-daemon at llvm.org (bugzilla-daemon at llvm.org)
Date: Mon, 02 Jan 2012 22:02:54 +0000
Subject: [LLVMbugs] [Bug 11692] New: An alloca that is never used for
anything but updating itself is not optimized away
Message-ID:
http://llvm.org/bugs/show_bug.cgi?id=11692
Bug #: 11692
Summary: An alloca that is never used for anything but updating
itself is not optimized away
Product: new-bugs
Version: 3.0
Platform: PC
OS/Version: Linux
Status: NEW
Keywords: code-quality
Severity: enhancement
Priority: P
Component: new bugs
AssignedTo: unassignedbugs at nondot.org
ReportedBy: nodam at otinn.com
CC: llvmbugs at cs.uiuc.edu
Classification: Unclassified
Created attachment 7829
--> http://llvm.org/bugs/attachment.cgi?id=7829
example showing the behaviour
An alloca that is never used for anything but updating itself is not optimized
away in the attached example. This depends on the size of the array. When I
tested it then it was optimized away if the size of the array was <= 21 and not
optimized away otherwise.
--
Configure bugmail: http://llvm.org/bugs/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are on the CC list for the bug.
From bugzilla-daemon at llvm.org Mon Jan 2 20:47:31 2012
From: bugzilla-daemon at llvm.org (bugzilla-daemon at llvm.org)
Date: Tue, 03 Jan 2012 02:47:31 +0000
Subject: [LLVMbugs] [Bug 11685] clang ignores -funsafe-math-optimizations,
-ffast-math, etc
In-Reply-To:
References:
Message-ID:
http://llvm.org/bugs/show_bug.cgi?id=11685
Chandler Carruth changed:
What |Removed |Added
----------------------------------------------------------------------------
Status|NEW |RESOLVED
CC| |chandlerc at gmail.com
Resolution| |FIXED
--- Comment #1 from Chandler Carruth 2012-01-02 20:47:31 CST ---
Fixed in r147434 and r147440.
--
Configure bugmail: http://llvm.org/bugs/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are on the CC list for the bug.
From bugzilla-daemon at llvm.org Tue Jan 3 00:10:08 2012
From: bugzilla-daemon at llvm.org (bugzilla-daemon at llvm.org)
Date: Tue, 03 Jan 2012 06:10:08 +0000
Subject: [LLVMbugs] [Bug 11690] Support building chromium in shared_library
mode
In-Reply-To:
References:
Message-ID:
http://llvm.org/bugs/show_bug.cgi?id=11690
Rafael ?vila de Esp?ndola changed:
What |Removed |Added
----------------------------------------------------------------------------
Status|NEW |RESOLVED
Resolution| |FIXED
--- Comment #3 from Rafael ?vila de Esp?ndola 2012-01-03 00:10:08 CST ---
Fixed in r 147441.
--
Configure bugmail: http://llvm.org/bugs/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are on the CC list for the bug.
From bugzilla-daemon at llvm.org Tue Jan 3 05:40:00 2012
From: bugzilla-daemon at llvm.org (bugzilla-daemon at llvm.org)
Date: Tue, 03 Jan 2012 11:40:00 +0000
Subject: [LLVMbugs] [Bug 11693] New: cannot install without permissions to
write to source tree
Message-ID:
http://llvm.org/bugs/show_bug.cgi?id=11693
Bug #: 11693
Summary: cannot install without permissions to write to source
tree
Product: new-bugs
Version: 3.0
Platform: PC
OS/Version: Linux
Status: NEW
Severity: release blocker
Priority: P
Component: new bugs
AssignedTo: unassignedbugs at nondot.org
ReportedBy: devel at fresse.org
CC: llvmbugs at cs.uiuc.edu
Classification: Unclassified
When issuing make install as a privileged user, the installation script tries
to write to the source directory.
freundt at segen:pts/19:/home/src/devel/llvm-3.0.src> sudo make V=1 install
...
llvm[1]: Regenerating LibDeps.txt.tmp
/bin/sh: /home/src/devel/llvm-3.0.src/tools/llvm-config/LibDeps.txt.tmp:
Permission denied
make[1]: *** [/home/src/devel/llvm-3.0.src/tools/llvm-config/LibDeps.txt.tmp]
Error 1
make[1]: Leaving directory `/home/src/devel/llvm-3.0.src/tools/llvm-config'
make: *** [install] Error 1
Reproducible: always
How to reproduce:
1. run ./configure and make as a user that is allowed to write to the source
tree
2. run make install as a user that isn't allowed to write to the source tree
In my case the user root is allowed to write to /usr but NOT to the source tree
itself.
--
Configure bugmail: http://llvm.org/bugs/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are on the CC list for the bug.
From bugzilla-daemon at llvm.org Tue Jan 3 06:28:37 2012
From: bugzilla-daemon at llvm.org (bugzilla-daemon at llvm.org)
Date: Tue, 03 Jan 2012 12:28:37 +0000
Subject: [LLVMbugs] [Bug 11694] New: FHS-ize info pages installation path
Message-ID:
http://llvm.org/bugs/show_bug.cgi?id=11694
Bug #: 11694
Summary: FHS-ize info pages installation path
Product: Build scripts
Version: 3.0
Platform: PC
OS/Version: Linux
Status: NEW
Severity: enhancement
Priority: P
Component: Makefiles
AssignedTo: unassignedbugs at nondot.org
ReportedBy: marius.maraloi at gmail.com
CC: llvmbugs at cs.uiuc.edu
Classification: Unclassified
To better adhere to the FHS-2.3 recommendations, we can update the path for the
installed info files. I'm proposing the following modification of "configure"
and "autoconf/configure.ac" scripts:
eval LLVM_ETCDIR="${prefix}/etc/llvm";
eval LLVM_INCLUDEDIR="${prefix}/include";
-eval LLVM_INFODIR="${prefix}/info";
+eval LLVM_INFODIR="${prefix}/share/info";
eval LLVM_MANDIR="${prefix}/man";
LLVM_CONFIGTIME=`date`
As most distrbutions come with at least a symlink between "/usr/info" and
"/usr/share/info", I'm submitting this as "enhancement" severity.
--
Configure bugmail: http://llvm.org/bugs/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are on the CC list for the bug.
From bugzilla-daemon at llvm.org Tue Jan 3 11:08:41 2012
From: bugzilla-daemon at llvm.org (bugzilla-daemon at llvm.org)
Date: Tue, 03 Jan 2012 17:08:41 +0000
Subject: [LLVMbugs] [Bug 11695] New: clang++ DWARF DW_TAG_friend generated
wrong
Message-ID:
http://llvm.org/bugs/show_bug.cgi?id=11695
Bug #: 11695
Summary: clang++ DWARF DW_TAG_friend generated wrong
Product: new-bugs
Version: 2.9
Platform: PC
OS/Version: Linux
Status: NEW
Severity: normal
Priority: P
Component: new bugs
AssignedTo: unassignedbugs at nondot.org
ReportedBy: davea42 at gmail.com
CC: llvmbugs at cs.uiuc.edu
Classification: Unclassified
Created attachment 7832
--> http://llvm.org/bugs/attachment.cgi?id=7832
source code used to generate bad DWARF
DW_TAG_friend is first documented in DWARF4 and should have
a DW_AT_friend attribute, not the attributes listed here
(attributes as shown by dwarfdump -i -M).
< 4><0x00000dd4> DW_TAG_friend
DW_AT_type <0x00000272>