similar to: [LLVMdev] make builds Release configuration by default

Hi! I'm trying to run "llc" tool on files from "test" subdirectory of LLVM root. I compile source files with the following line: llvm-gcc -c -ocompiled.o ackermann.c Then I run llc like the following: llc compiled.o which outputs llc: bytecode didn't read correctly (I mean, I try it with different options but even the default ones aint working) My llvm-gcc

Hi! I've did what Jim Laskey wrote but llc didn't reckognize my regalloc option. So I moved my allocator implementation into seperate folder within CodeGen and wrote separate makefile for it (like in "Writing an LLVM pass" tutorial). But when I run "make" from LLVMOBJDIR it doesn't enter the RegAlloc directory and when linking llc an error like

Dear Anton, you can add your register allocator strait iin the "lib/CodeGen/Passes.cpp", and then 're-make' it: "makellvm llc", on the top of lib/CodeGen. It is faster than running make from LLVMOBJDIR. The problem is that it only add to llc the changes on the lib/CodeGen directory. If you change other parts, a make from LLVMOBJDIR will synchronize it. Try adding

I managed to link my RegAlloc.a library to llc tool but can;t make the same with lli tool. Should I use USEDLIBS variable in lli makefile (like llc) or there is another way? On 8/20/06, Fernando Magno Quintao Pereira <fernando at cs.ucla.edu> wrote: > > > Dear Anton, > > you can add your register allocator strait iin the > "lib/CodeGen/Passes.cpp", and then

So what addIntervalsToSpills returns are new intervals to allocate with infinite weights, right? And I need not to allocate the old interval. Should hasStackSlot return true on its register then? On 8/21/06, Fernando Magno Quintao Pereira <fernando at cs.ucla.edu> wrote: > > > Well, someone correct me if am wrong, but, you still have to allocate > physical registers to them,

On Thu, 3 Aug 2006, Jim Laskey wrote: > To force the load/linking of your register allocator into the llc/lli tools, > add your create function's global declaration to "Passes.h" and add a > "pseudo" call line to "llvm/Codegen/LinkAllCodegenComponents.h" . Another note: with this new functionality you should be able to dynamically load register

Tony, By default, llvm-gcc4 does not produce bytecode, it produces a native object file, just like GCC does. So, if you try to pass it to llc, it is obviously going to report problems with the bytecode. Either skip the llc step or use the -emit-llvm option: llvm-gcc -c -emit-llvm -o compiled.o ackermann.c llc compiled.o -o compiled.s Reid. On Sun, 2006-08-20 at 20:26 +0400, Anton Vayvod wrote:

Hi! I developed a register allocator within LLVM and now I need to test its efficiency. Can I do this using llvm-test package? Do llvm tests check all available regalloc options automatically? If not, then what modifications should I do to the test files? It would be great if I could test my algo along with linearscan and compare the results. Thanks. Tony. -- "Nae king! Nae quin! Nae

Well, someone correct me if am wrong, but, you still have to allocate physical registers to them, because their values must be reloaded before been used. But after you spill a register, you break its live ranges, so, each use of that register can be loaded into a different physical. In my register allocator, after I spill reg_v: For each use of reg_v: create a new reg_v' replace reg_v

The current CVS version doesn't compile on Linux: MachOWriter.h:583: error: 'strlcpy' was not declared in this scope Rafael

On 8/7/06, Chris Lattner <sabre at nondot.org> wrote: > > On Sun, 6 Aug 2006, Anton Vayvod wrote: > > I'm developing a register allocator that works iteratively. It spills > some > > virtual registers on each iteration until all the rest have physical > ones > > assigned. > > Take a look at the linear scan allocator. It is also iterative: it uses >

I'm not sure about one thing: you assign stack slot to each new register you replace the spilled one with. And then you need to allocate physical registers to them. Is it possible to assign physical register to the virtual one which has a stack slot already? On 8/21/06, Fernando Magno Quintao Pereira <fernando at cs.ucla.edu> wrote: > > > > So what addIntervalsToSpills

> So what addIntervalsToSpills returns are new intervals to allocate with > infinite weights, right? > And I need not to allocate the old interval. Should hasStackSlot return true > on its register then? > I am not very sure about addIntervalsToSpill, but, for all the registers created to replace a spilled registers, they must have a stack slot assigned to them. I am sending you my

Hi guys I have a reasonably basic question with zoo usage, but I havent been able to find a satisfactory workaround yet. Heres a simple example of what I'm talking about (the following data frame contains numeric columns that contains NAs): > head(ebs) time src tstamp code bid ask 1 2009-03-03 13:03:29.536 perf.Tib_listener 14980321164 EBS.REC.EURJPY=EBS.NaE 123.48 NA 2 2009-03-03

On Tue, 2007-10-02 at 14:38 -0700, Tanya M. Lattner wrote: > > I've read and followed the README.LLVM file that comes with LLVM-GCC 4.0 > > Front End source, but got the following error when compiling: > > > > configure: error: You must specify valid path to your LLVM tree with > > --enable-llvm=DIR > > make: *** [configure-gcc] Error 1 > > > >

Greetings I'm trying to build LLVM 2.6 GCC from the source files, but I am having problems during the make step. Here is the information: 1. My machine is: Linux firefly 2.6.9-89.0.9.ELsmp #1 SMP Wed Aug 19 08:06:10 EDT 2009 x86_64 x86_64 x86_64 GNU/Linux my gcc version is 4.2.1 2. I've unpacked llvm-gcc-4.2-2.6.source.tar.gz under the llvm-gcc directory. I've also created two

> I've read and followed the README.LLVM file that comes with LLVM-GCC 4.0 > Front End source, but got the following error when compiling: > > configure: error: You must specify valid path to your LLVM tree with > --enable-llvm=DIR > make: *** [configure-gcc] Error 1 > > I've done the following in csh before hand: > $ setenv LLVMOBJDIR

Hi Eliseu, > 2. I've unpacked llvm-gcc-4.2-2.6.source.tar.gz under the llvm-gcc > directory. I've also created two subdirectories > called "install" and "obj" under the same llvm-gcc. > > 3. Then I did the following, based on README.LLVM : > > %setenv LLVMOBJDIR /home/eliseuf/llvm/llvm-gcc/obj first you need to build llvm itself. LLVMOBJDIR

I've read and followed the README.LLVM file that comes with LLVM-GCC 4.0 Front End source, but got the following error when compiling: configure: error: You must specify valid path to your LLVM tree with --enable-llvm=DIR make: *** [configure-gcc] Error 1 I've done the following in csh before hand: $ setenv LLVMOBJDIR /home/napi/proj/c2jvm/llvm/llvm-gcc/obj Where did I go wrong?

If llvm is build with srcdir != objdir, building gcc will fail with ../../llvm-gcc4-1.7.source/gcc/llvm-internal.h:36:36: error: llvm/Support/DataTypes.h: No such file or directory This happens because DataTypes.h is a generated header. Adding -I$(LLVMOBJDIR)/include to INCLUDES in Makefile.in:878 solves the problem. One more thing, is the modified GCC available in some svn or cvs repository?