similar to: [LLVMdev] Adding "S" suffixed ARM/Thumb2 instructions

On Feb 17, 2011, at 10:35 PM, Вадим Марковцев wrote: > Hello everyone, > > I've added the "S" suffixed versions of ARM and Thumb2 instructions to tablegen. Those are, for example, "movs" or "muls". > Of course, some instructions have already had their twins, such as add/adds, and I leaved them untouched. Adding separate "s" instructions is

On Oct 14, 2013, at 3:27 AM, Zakk <zakk0610 at gmail.com> wrote: > Hi all, > I meet this problem when compiling the TREAM benchmark (http://www.cs.virginia.edu/stream/FTP/Code/) with enable-misched > > The small function will be scheduled as good code, but if opt inline this function, the inline part will be scheduled as bad code. A bug for this is welcome. Pretty soon, I’ll

Hi all, I meet this problem when compiling the TREAM benchmark ( http://www.cs.virginia.edu/stream/FTP/Code/) with enable-misched The small function will be scheduled as good code, but if opt inline this function, the inline part will be scheduled as bad code. so I rewrite a simple code as attached link (foo.c), and compiled with two different methods: *method A:* *$clang -O3 foo.c -static -S

Hi folks, me again, So, I want to implement a simple optimization in a NEON case I've seen these days, most as a matter of exercise, but it also simplifies (just a bit) the code generated. The case is simple: uint32x2_t x, res; res = vceq_u32(x, vcreate_u32(0)); This will generate the following code: ; zero d16 vmov.i32 d16, #0x0 ; load a

When the IR specifies a 32 bit load can it be changed to a narrower load? What if the load is from memory (e.g. a peripheral) that only supports 32-bit access? Consider the following IR: ---- target datalayout = "e-p:32:32:32-i1:8:8-i8:8:8-i16:16:16-i32:32:32-i64:32:32" target triple = "thumbv7m-unknown-unknown" @f = external global i32 define zeroext i8 @bar() nounwind { L.0:

I used this service several months ago and was very pleased with the response. I have a dataframe with several thousand lines and to each line I need to apply a series of "if else" statements. For each row I need either a value or a blank/NA. Below is the series of if else statements I have been trying without success to integrate into a function such as "apply". if

Hi Piotr, I think I found a working matcher: match ifStmt(hasCondition(implicitCastExpr(hasImplicitDestinationType(isInteger()), has(cxxMemberCallExpr(callee(cxxMethodDecl(hasName("compare"))), hasArgument(0, declRefExpr().bind("str2")), callee(memberExpr(has(declRefExpr().bind("str1"))))))))).bind("case1") This one bind to both str1 and str2 in

Hi Piotr, Thanks. Yeah, it seemed a little weird, but it was what got me closest. I found out that the matcher I supplied here was working for clang-query 3.8.1. I'm working on a clang-tidy module for 4.0.0 - it's not working there. Could you elaborate on the "onImplicitObjectArgument"? There is no document on it on the clang page. So I wouldn't know how it works or what it

Adding cfe-dev, because it is related to clang, not LLVM. 2016-11-27 22:34 GMT+01:00 Piotr Padlewski <piotr.padlewski at gmail.com>: > Hi Mads, > Can you provide the code that you run clang-query on, or at least AST for > the fragment you want to match? > > Piotr > > 2016-11-26 22:27 GMT+01:00 Mads Ravn via llvm-dev <llvm-dev at lists.llvm.org > >: >

Yes, I agree with you -------------- next part -------------- An HTML attachment was scrubbed... URL: <http://lists.llvm.org/pipermail/llvm-dev/attachments/20080106/efb1c465/attachment.html>

Hey again) Now I have next code: ; ModuleID = 'sample.lz' @.str1 = internal global [8 x i8] c" world!\00" ; <[8 x i8]*> [#uses=1] @.str2 = internal global [8 x i8] c"hello, \00" ; <[8 x i8]*> [#uses=1] @.str7 = internal global [21 x i8] c"welcome to out hall!\00" ; <[21 x i8]*> [#uses=1] declare i32 @puts(i8*)

but why this code don't work: ; ModuleID = 'sample.lz' @.str1 = internal global [6 x i8] c"world\00" ; <[6 x i8]*> [#uses=1] @.str2 = internal global [7 x i8] c"hello \00" ; <[7 x i8]*> [#uses=1] @.str7 = internal global [7 x i8] c"father\00" ; <[7 x i8]*> [#uses=1] @.str8 = internal global [8 x i8]

Dear community, I faced with code which was generated by llvm, assembly instructions of that code is relying on 8-bytes alignment for structures on the stack. The part of Objective C code is following: -(void)getCharacters:(unichar *)unicode {     NSRange range;     range.location = 0;     range.length = [self length];     printf("%p, %p\n", &range.location, &range.length); And

I am trying to understand the hello word assember example. This is my version: %str1 = internal constant [13 x sbyte] c"Hello World\0a\00" declare int %printf(sbyte*, ...) implementation ; Functions: int %main() { %str2 = getelementptr [13 x sbyte]* %str1, long 0, long 0 call int(sbyte*, ...) *%printf(sbyte* %str2) ret int 0 } Why is getelementptr being given two "long

Hi everyone, happy new year. This note is to announce that support for PC relative reloc tags for movw/movt is nearing completion (hopefully <48hrs!). This work is is from Jan Voung, David Meyer and myself. Unfortunately, to test this change, we need to patch ARM/AsmParser to address http://llvm.org/bugs/show_bug.cgi?id=8721 Locally, we have hacked up a solution to 8721, but its not ideal

Hi, Hope this is the right channel for this question. I am trying to make an ast matcher for clang-tidy to find str1.compare(str2), where both str1 and str2 are std::string. I have this so far: http://i.imgur.com/sUma9WC.png . But I am having a hard time finding a way to bind an id to the str1 part of the expression. Can anyone help me out with an idea? Best regards, Mads Ravn --------------

Hi, I have a vector, which contains both strings and numbers, e.g. > foo <- c("str1",1234,"str2",0.9876) I want to know if a distinct element of the vector is a string or a number and took "is.numeric", but > is.numeric(foo[2]) [1] FALSE because R treats the numbers in a mixed vectors as strings: > foo [1] "str1" "1234"

-llvmcommits On Mon, Jan 10, 2011 at 3:21 PM, Renato Golin <renato.golin at arm.com> wrote: > Btw, I know this is for ELF printing, but can the same infrastructure > you're using to print the hi/lo be used to print relocation in Asm > output? Or is this a completely separate subject? Hi Renato, If I am understanding you correctly, then the answer is no, because .s output

I have some elements in a vector with extraneous information (e.g. file name and sample IDs) that I'd like to strip from every element. For example, I would like "SPI1.S1.str1.P3.sample.tif" "SPI1.S1.STR2.P1.sample.tif" to read "SPI1.S1.str1.P3" "SPI1.S1.STR2.P1". Will someone help me with the syntax in gsub? It needs to be something like

Dear developers: When I am doing basicblock pass, I meet a bug: there is an iterator "I" in a basicblock, and it is not pointing to the first instruction in this basicblock. However, "I--;" will fail by an assertion. The basic block ("I" is pointing to the second instruction) in test.ll: bb: ; preds = %bb1 %1 = call i32 (i8*, ...)* @printf(i8* noalias