similar to: profiles, profdata & homes

Duncan, Justin, I'm about to submit a series of patches that add writing capabilities for sample profiles in both text and binary formats. Soon, I'll add a third format (to make it interoperable with GCC). I would like to add some profile maintenance utilities as well: merging, dumping and converting. It seems like the best place would be tools/llvm-profdata. But that means that I need

Hi! UBSan bootstrap bot fails with error report on 5 llvm-cov and llvm-profdata lit-tests: http://lab.llvm.org:8011/builders/sanitizer-x86_64-linux-bootstrap/builds/4526/steps/check-llvm%20ubsan/logs/stdio Also see http://llvm.org/bugs/show_bug.cgi?id=20815. The error reports look vaild: in general it's incorrect to load uint64_t, or even structures like "RawHeader" or

I haven't tested it, but it looks to me like llvm-profdata merge (well, InstrProfWriter specifically) would not have deterministic output. Certainly the textual output iterates over FunctionData which is a StringMap of SmallDenseMaps, neither of which has deterministic iteration. The binary writing looks like it'd have similar issues - looping through these unordered maps & writing

I'm trying to set up for my small school lab: Samba as PDC, 11 workstations running XP Pro, roaming profiles with folder redirection. (I finally got the workstations to join the domain by adding them by hand--so I assume something is wrong with my add machine script, but I'll address that issue later, since I have only the 11 XP boxes to deal with.) Now I'm working on getting the

Yes, that was it! Now that I took a closer look, the guide also states that I should use the stage2 build. Silly me. Thanks! On Tue, 3 Sep 2019 at 19:31, David Blaikie <dblaikie at gmail.com> wrote: > I /guess/ you actually want /path/to/release_build/llvm-profdata because > the profiles are generated from binaries compiled with the release build, > so it's the release build

Hi! I'm trying to build a fast Clang for myself to use for debug builds on Clang itself, but I've been struggling for a very long time on it. Could you please help? I've been following this guide: https://llvm.org/docs/HowToBuildWithPGO.html I've quickly learned that its outdated, because the script it talks about doesn't work with the monorepo layout at all, but in any

Hi all, I'd liked to get your thoughts on possibly adding a generic key-value store to the profile data formats for 'metadata'. Some potential uses cases: *I. Profile Features* The most basic use could be as a central repository for internal bits of housekeeping information about the profile data. For example, to differentiate between FE and IR instrumentation:

On Thu, Jun 29, 2017 at 7:03 PM Xinliang David Li <davidxl at google.com> wrote: > On Thu, Jun 29, 2017 at 6:27 PM, David Blaikie <dblaikie at gmail.com> wrote: > >> I haven't tested it, but it looks to me like llvm-profdata merge (well, >> InstrProfWriter specifically) would not have deterministic output. >> >> Certainly the textual output iterates

On Fri, Jan 15, 2016 at 11:41 AM, Xinliang David Li <davidxl at google.com> wrote: > Tagging profile data with such information is generally useful. My > thoughts are > > 1) such information is probably not needed to be stored in raw format > profile data -- so no runtime changes are needed -- only llvm-profdata > and indexed format need to be enhanced to support this.

On Thu, Jun 29, 2017 at 7:26 PM, David Blaikie <dblaikie at gmail.com> wrote: > > > On Thu, Jun 29, 2017 at 7:03 PM Xinliang David Li <davidxl at google.com> > wrote: > >> On Thu, Jun 29, 2017 at 6:27 PM, David Blaikie <dblaikie at gmail.com> >> wrote: >> >>> I haven't tested it, but it looks to me like llvm-profdata merge (well,

Hi all! It seems I came across on issue with coverage mapping (http://www.llvm.org/docs/CoverageMappingFormat.html) check on: llvm revision: r228136 clang Last Changed Rev: 228121 build: Debug+Asserts OS: ubuntu 14.04 Here is simple snippets test1.c: NOT OK ================== #include <stdio.h> static int foo() { return 42; } int main() { return 0; } ================== cp src/test1.c

Dear List, This is my first post here, but yesterday I finally configured a full PDC with LDAP backend using Chapters 4 and 5 of the Samba Guide (exact same directory and shares setups just now). Very well written! I am slowly getting the other things like printing and DNS added, but I wanted to confirm/ask a few things. When I first logged on I got complaints about not being able to create the

Hi Diego, thanks for clarifying the difference between the two formats. I have noticed the new note in the "Sample Profile Format" section of the Clang guide clarifying that it is different from the coverage format. So, my further question is... Am I right in understanding that both formats can be used for PGO purposes then? I have tried the following, as in the Clang user guide: $

Hello all, I'm trying to get folder redirection working with roaming profiles and Samba 3.0.2a and Windows XP clients. I've followed the Samba 3 by Example document and have change the User Shell Folders registry key and created a Default User profile in the netlogon share and set the machnes group policy for excluding the directories from the roaming profile. Everything is working

I would like to setup a samba PDC with mandatory profiles for my school. I know that I can easily use them for XP (centralized) and that for Win 98 I have to put them in the user accounts. Now my question: do I have to put a profile in the user account each time a new user is made? Or is there a simple way of doing this. (off course I would link it to one profile, but this way I still have to

This week I migrated out main server from Fedora 8 to Fedora 11. It has been a stressful time, but things are mostly working. Samba and LDAP weren't really a problem, but I've beat down the list of problems to something Samba related. (For the benefit of anyone else going this route, the biggest problem by far was iptables. Maybe we had it turned off under FC8, but I suspect

Hi all, I am working on adding PGO to LDC (LLVM D Compiler). The current implementation 1) uses LLVM's InstrProf pass to generate an instrumented executable 2) links to compiler-rt/lib/profile for the runtime functionality to write a raw profile data file 3) uses llvm-profdata to merge profile data and convert from profraw to profdata format 4) uses llvm::IndexedInstrProfReader to read-in

We are trying to get PGO working for our embedded out-of-tree target, but the utility 'llvm-profdata' does not like the data we are giving it. Because this is not a hosted environment, we have to off-chip the profiling data ourselves, and although the data looks okay, 'llvm-profdata' reports the following error: llvm-profdata show -all-functions -counts -detailed-summary -text

> > I think it is reasonable to simply replace the key we currently use with > MD5(key) for getting a size reduction. In practice for my use cases, I have > not observed any of the issues you mentioned under "Large size of overhead > can limit the usability of PGO greatly", but I can understand that some of > these issues could become problems in Google's use case.

Hi I encountered this problem a few times in the past, but only now it is consistent enough so I can write this email. In short, the problem is that after I copy some structure from the profiled program to the dtrace space, and set a "this-dataP" variable to point to the local copy, after a while some of the fields of the local structure are overridden with junk values. In