similar to: Google Season of Docs

Hey everyone, I've made several updates to the LLVM docs site over the last few weeks. I'd like to get some feedback before I make any additional changes. As of right now, here are the updates I've made: a.) Added a static sidebar with links to the docs, community pages, etc. b.) Moved the Community/Getting Involved info to a separate page. c.) Consolidate the docs into three

hi, Here is wishing everyone in the CentOS community a great holiday season, awesome new year and a great 2014. May 2014 bring friendlier + larger communities, stabler distros and a general reduction in systems frustrations for the users! regards from and to everyone, - KB -- Karanbir Singh +44-207-0999389 | http://www.karan.org/ | twitter.com/kbsingh GnuPG Key :

Hi David, Thanks for your reply. I find that I need to add some new types of profile data that are specific to JIT environment like Function Ordering. Function Ordering is similar to dynamic call graph which records the execution of functions at runtime along with the order in which they are called. Eg: Suppose they are 5 functions (F1..F5). F1 calls other functions in the order described

I'm still relatively new to llvm, good to know these errors are from CodeGen. On 03/03/19 10:49 AM, Tim Northover wrote: > On Sat, 2 Mar 2019 at 21:00, preejackie via llvm-dev > <llvm-dev at lists.llvm.org> wrote: >> Could you please explain how you are using this? like commands >> >> When I run this IR in opt, it doesn't throw any errors. > The error

Hi David! Thanks for reaching out, the codegen part of the tutorial is fine, but when I try to generate the object code file, it throws segfault. Also please note that I'm using release build of llvm, actually I don't know whether that is the root cause or not. I also run through valgrind, it seems like a invalid read from stack. |||Error Summary: | |1 errors in context 1 of 1:

Hi, David: Good point, it will be interesting to see speculative compilation in this context applying on devirtualization, with high level (type) information if applicable. > On Mar 28, 2019, at 2:35 PM, preejackie <praveenvelliengiri at gmail.com> wrote: > > Hi David & Bekket, > > Thanks your replies :) > > David, Indent here is: ORC v2 JIT APIs has

Hi all, I'm looking for some program analysis techniques which help me to find potential functions to execute next, from the current executing function. I want to decision based on compile time information. I consider LLVM IR is too low-level to make such analysis. So, I using call graph representation of module. I figured out the probability of function which execute next based on the

Hi all, I need to do dynamic profiling in ORC JIT between the runs. Is it possible to re-use parts of the static pgo code in llvm for the purpose given that JIT uses llvm codegen. Or I need to write my own implementation to support profile guide optimization in JIT ? Please tell me if you want information. -- Have a great day! PreeJackie -------------- next part -------------- An HTML

Devirtualization is an example of predicting calls and is much more easily done on a higher-level representation. It is simply easier to reason about certain things given information that is lost during translation to LLVM IR. The MLIR project makes similar arguments. It would be helpful to know what's being attempted here. I'm not sure what the (hardware?) branch predictor has to do

Hi, Could you please explain how you are using this? like commands When I run this IR in opt, it doesn't throw any errors. On 03/03/19 1:30 AM, via llvm-dev wrote: > Hi, > > I'm having an error I've never seen before > > LLVM ERROR: Cannot select: t20: ch,glue = <<Unknown Target Node #284>> t11, TargetGlobalAddress:i32<i32 (i32, i32)* @add2> 0,

Hi Bekket, Thank you for your reply. Earlier I came across a paper called "Dynamic Look Ahead Compilation: Hide JIT compilation latencies", it devised some methods for JIT compilation of functions before the actual call takes place by using call graph analysis, branch probabilities to generate a list of functions with high likelihood of execution in near future. In my opinion it

Heyho, I can only provide you with a screenshot for the first problem I noticed: [https://media.discordapp.net/attachments/534012750045642783/575369402195640321/unknown.png] This code will crash when calling "eeBuilder.create". But I also looked at what happens if I do reuse the llvm::ExecutionEngine. After the call to "finalizeObject" I use "removeModule" for the

Hi Bekket, Sorry for the delayed reply. By appropriate, I mean performance of compiled native code. I was referring other JIT implementations like LuaJIT & webkit FTL JIT to see how they implement their JIT. I have gone through the design of Spider Monkey & Android runtime (ART) JIT. As, you said both region based and method based compilation unit have their own advantage &

Hello everybody, I need to calculate seasonal means with temperature data for my work. I have 70 files coming from weather stations, which looks like this for example: startdate <- as.POSIXct("01/01/2006", format = "%d/%m/%Y") enddate <- as.POSIXct("05/01/2006", format = "%d/%m/%Y") date <- seq(from = startdate, to = enddate, by =

preejackie via llvm-dev <llvm-dev at lists.llvm.org> writes: > Are you suggesting that static analysis is much inferior than > profiling in finding "next executing function" ? I'm stuck at > deciding which one to prefer & implement during this summer project, > Could you please help me? I suspect this is highly code-dependent. For example, it's probably

Hi all, I am trying to interparete the result of the following output from lm; fit1 =lm(Feed _Intake ~ weight + season + weight*season) Season has three classes(x,y,z) Reults are Estimate (Intercept) 21.51559 weight 2.13051 factor(season)y 10.59739 factor(season)z 1.30421 weight:factor(season)y

On Thu, May 03, 2018 at 01:09:30AM -0400, Kevin Korb via rsync wrote: > --delete only works when you sync a dir not individual files. So... > rsync -vai --delete --include='g*' --include='G*' --exclude='*' > /share/Public/videos/ /share/Backup-HDD-03 That kinda didn't work: rsync -vai --delete --include='S[A-Z]*' --include='S[a-h]*'

Hi all, I'm interested in participating this GSoC with llvm. I would like to contribute on the project idea "DebugInfo should not effect codegen". Over the past few days, I have gone through some bugs reported at bugzilla that cause different codegen behavior of same program. Bugs I have been looking at are:     1. [fuzzDI] -O1 + `-g` cause the generated code to change.

Dear R list, I have a sequence of weekly observations of number of adults and larvae in various size classes from a butterfly population living in a subtropical area with pronounced wet and dry seasons. Wet and dry seasons are each defined 26 weeks long with fixed start and end dates. The data span 103 weeks (two seasons each of wet and dry) with some missing weeks. What I would like to do is

Hi all I can't find what I am looking for so I am asking here. I have a dataset that looks something like this. Year season percent_below 2000 Winter 6.9179870 2000 Spring 1.6829436 2000 Summer 1.8463501 2000 Autumn 3.8184993 2001 Winter 2.8832806 2001 Spring 2.5870511 2001 Summer 0.0000000 2001 Autumn 4.7248240 2002 Winter