similar to: LLVM Optimizations strange behavior/bug

Thanks a lot for your help. Indeed if I compile with -OZ the loop gets replaced with something like this: if ( *v3 < 0 ) v4 = 0; Is there a reason why it does not get applied with O3 ? I mean this a huge improvement to the loop. -----Original Message----- From: Doerfert, Johannes [mailto:jdoerfert at anl.gov] Sent: Dienstag, 13. August 2019 19:13 To: llvm-dev at lists.llvm.org; Garba

Hi, I was wondering if there is an optimization pass in LLVM or Polly that can optimize this given loop: int main(int argc, char* argv[]) { char s = 0; // Loop is setting s = argv[1][0]; for (int i=0; i<argv[1][0]; i++) { s++; } printf("s = %c\n", s); return 0; }

Hello, I am experiencing problems, when trying to assemble these two x86-64 Opcodes "add r64, imm32" "imul r64, r64, imm32" When having the most significant bit set for imm32, for example: "add rax, 0x80000000", "add rax, 0xffffffff", ... "imul rbx, rsi, 0x80000000", "imul rbx, rsi, 0xffffffff", ... The Error Message I receive is the

Hi All, I tryed to create a global structured derived declaration in my compiler by: StructType *st = StructType(getGlobalContext(), elements, false); In dump module i didn't find any code needed to declare a new user data type. Is there a instruction to add this class into current module? Cheers, Manuele

I''ve upgraded from 1.9.0 to 1.9.2 and am finding that the errors I''m getting don''t match the docs. For example, in Wx::ListCtrl: - The doc makes no mention of on_get_item_column_image(), but without this method I get NoMethodError. - The doc says on_get_item_attr() *may* be overloaded, but if I don''t define it, I get NoMethodError. The API doc appears to

Hi, I try to understand why llvm can't optimize this small piece of code. My expectation is that the IR will be optimized to a " ret i64 2101545. But somehow the alias analysis fails when the pointer is casted to an integer value, so the dead code eliminator can't optimize it away. Could someone explain to me why this is failing and if there is a solution for that ? I can trick the

I''m getting the following errors multiple times per request when using 4.3.0. I do not receive any errors when using 4.2.1. Please CC me on replies, I''m not subscribed to the mailing list. 16:48:42 web.1 | E, [2012-04-26T16:48:42.733954 #87940] ERROR -- : app error: Socket is not connected (Errno::ENOTCONN) 16:48:42 web.1 | E, [2012-04-26T16:48:42.734193 #87940] ERROR

I''m expermienting with Shorewall for use protecting my servers(about 25 all currently using pub addresses). My pointy-haired-boss is somewhat sceptical about adding a firewall that would be a single point of failier for our services. Does anyone have a good set of documentation that I could use in setting up either an Active/Active or Active/passive setup with two identical shorewall

-----BEGIN PGP SIGNED MESSAGE----- Hash: SHA1 Try it without any --delete options. On 03/27/2015 09:31 AM, Aron Rotteveel wrote: > I am now running with --delete --numeric-ids --relative but the > problem still persists. > > -- Best regards / Met vriendelijke groet, > > Aron Rotteveel > > 2015-03-27 14:22 GMT+01:00 Kevin Korb <kmk at sanitarium.net >

I''d like to see if we can use DTrace to as the kernel implementation of the BSD systrace security policy system (http://www.systrace.org). I don''t really want to port systrace to Solaris because I think with DTrace we already have all the necessary in kernel hooks to do this. With systrace you express things like: "httpd can bind to port 80 but not any other port, it

Hello Gene, I read and nodded sometimes and caught myself thinking that I used to propose or agree with some of the points you make, but outgrew those. > ...makers claim... I suppose this means asciidoc makers (not e.g. UPS makers)? They are actually largely non-Windows, especially in the part about rendering books, presentations, local HTML pages or web sites. These are all

Hi, I test your code and it works!!! but I didn't understand if "create a function" it necessary for my aims. What I tried to do is declare a structured to allow after declaration global or local structured variables. How I can I do? Thanks, Manuele Il 31/01/2013 10:14, Garba,Peter ha scritto: > Hi, > > that's what I'm doing ... > > //Init the struct type

Dear LLVM development team, working with the LLVM gold plugin, I have encountered an unexpected behaviour when the option --allow-multiple-definition (or -z muldefs) is specified for the linker. Let's suppose the following scenario with four simple source files: ----- main.c ------ #include "unit.h" int main() { only_in_unit1(); only_in_unit2(); return

-----BEGIN PGP SIGNED MESSAGE----- Hash: SHA1 Try also removing --delete-excluded. Without those two options there should be no reason for rsync to require gigs of RAM. Well, unless the other system has rsync 2.x. On 03/27/2015 07:29 AM, Aron Rotteveel wrote: > Yes, I removed "--no-inc-recursive", without success. > > -- Best regards / Met vriendelijke groet, > >

Hi all, A couple of people are trying to convince me to come to FOSDEM in Belgium at the end of February. I''m a bit hesitant because I''ve been doing so much travel, but LinuxConf.au was a great experience and seems to have really helped get exposure to Puppet. There seems to be a pretty strong European contingent of Puppet users, and it''d be great to meet

Hi, I've setup a Samba 4 based AD server a while ago and everything was working fine for some time. However, since a few days it seems it doesn't resolve aliases anymore. I can query them using the host command, but a simple ping then fails to resolve the name: % host alias1 alias1.altum.de is an alias for server1. % ping alias1

For a company study I am trying to come up with what would one need to replicate *grosso modo and at the functional level* what we get in Rails. This does not need to be an "equivalent" solution, just what a Java programmer would build upon, the canonical aggregation of components, the usual tools. My first fraft is: Struts -> C layer Hibernate -> M

Hi Rusty, I had a look over your 011-paravirt-head.S.patch. I'm struggling to come up with a list of any benefits over having separate entrypoints for each hypervisor. Multiple entry pros: * allows maximum startup flexibility for any given hypervisor * for Xen at least, it's pretty simple * the "what hypervisor am I under" question is answered trivially cons:

https://bugs.freedesktop.org/show_bug.cgi?id=100758 Bug ID: 100758 Summary: Kernel Oops when starting X nv50_dma_push Product: xorg Version: unspecified Hardware: x86-64 (AMD64) OS: Linux (All) Status: NEW Severity: normal Priority: medium Component: Driver/nouveau Assignee:

Am 22.04.2017 um 14:56 schrieb Luca Olivetti via samba: > Check nsswitch.conf. Make sure you just have "iiles dns" in the hosts > line. Yes, that's the case. Bye... Dirk -- Dirk Heinrichs <dirk.heinrichs at altum.de> GPG Public Key: D01B367761B0F7CE6E6D81AAD5A2E54246986015 Sichere Internetkommunikation: http://www.retroshare.org Privacy Handbuch: