similar to: [LLVMdev] XOR Optimization

Hi, On Tue, Jul 26, 2011 at 11:32 AM, Matt Johnson <johnso87 at crhc.illinois.edu>wrote: > Hi Daniel, > > > Hi folks, > > > > I couldn't find a specific XOR (OR and AND) optimization on llvm, and > > therefore I am about to implement it. > > But first I would like to check with you guys that it really does not > exist. > > > > For a

Hi folks, I couldn't find a specific XOR (OR and AND) optimization on llvm, and therefore I am about to implement it. But first I would like to check with you guys that it really does not exist. For a simple loop like this: nbits = 128; bit_addr = 0; while(nbits--) { bindex=bit_addr>>5; /* Index is number /32 */ bitnumb=bit_addr % 32; /* Bit number in longword */

Hi- > I haven't seen a machine in which OR is faster than ADD nor more energy-efficient. They're all done by the same ALU circuitry which delays the pipeline by its worstcase path timing. So, for modern processor hardware purposes, OR is exactly equal ADD. Transforming ADD to OR isn't strenght reduction at all. Maybe this is benefical only if you have a backend generating circuitry

Den 26.07.2011 03:25, skrev Daniel Nicácio: > - Do you know why a OR instruction is used for increments? instead of > using a INC or ADD? You mean "bit_addr++;" ? It's add'ing (+4 due to loop unrolling i guess) %inc.3 = add i32 %0, 4

"The fact that the loop is unrolled explains why the XORs, SHLs, and ORs are not folded into 1." I dont see why the unrolling explains it. "I think he is trying to say this expression generated by unrolling by a factor of 4 can indeed be folded into a single XOR, SHL and OR. " Precisely. The code generated by unrolling can be folded into a single XOR and SHL. And even if it

Hi people Just switched to Dovecot yesterday, and I've already got about five reports of users with duplicate mail in their inbox :-/. Not as bad as lost email, mind you, but pretty annoying still. Checking the server log, it says: postfix/smtpd: disconnect from unknown deliver: msync() failed with index file maildir/dovecot.index: Input/output error unionfs: new lower inode mtime

Hi Duncan, when I run "opt -std-compile-opts" on the original source code it has the same output of O3. when I run "opt -std-compile-opts" on the -O3 optimized code, things get even more weird, it outputs the following code: while.body: ; preds = %while.body, %entry %indvar = phi i32 [ 0, %entry ], [ %indvar.next.3, %while.body ] %tmp

After a few more tests, I found out that if we set -unroll-threshold to a value large enough, and run "opt -std-compile-opts" or "opt -O3" 3 times, the unroll will be able to unroll the original loop 32 times, and when you have it unrolled for at least 32 times a optimization is triggered, folding it to a single "%xor.3.3.1 = xor i32 %tmp6, -1" (dont know why it does

Hi, After a few distractions I'm back to work on the llvm tgsi backend. I've added clang integration and I can now compile a simple opencl program to something which sort of looks like tgsi. You can find my latest work on this here: http://cgit.freedesktop.org/~jwrdegoede/llvm http://cgit.freedesktop.org/~jwrdegoede/clang (the latter may still need to sync) I've a little test

Hi Daniel, > Precisely. The code generated by unrolling can be folded into a single XOR and > SHL. And even if it was not inside a loop, it can still be optimized. What I > want to know is: is there any optimization supposed to optimize this code, but > for some reason it thinks it is not possible, or there is no optimization for > that situation at all? it could be a phase

So where does this leave things regards Eaton and shut down? Has anyone managed graceful shut down from an Eaton unit with either of the ushhid or bcmxcp drivers? Perhaps I should be coming at this from the other end of the stick. What manufactures have the best support in NUT that I can look at and be confident I'm likely to have a successful deployment? Really the only deal breaker

Hello, I'm trying to fill CID data into the astdb using AsteriskWin32's asterisk.exe, to no avail: The batch file stops after the first line, and just waits: ---------------------------------------- rem c:\cygroot\mystuff>import.bat rem rem c:\cygroot\mystuff>C:\cygroot\bin\asterisk.exe -rx 'database put cidname 123 "My cellphone"' rem rem Asterisk module

Dear R users, I'd like to make this data rem.y = c(-1,0,2,4,5) from y = c(-1,-1,0,2,2,2,2,4,4,5,5,5,5,5). That is, I need to remove repeated values. Here is my code, but I don't think it is efficient. How could I improve this? #------------------------------------------------------------------------ y = c(-1,-1,0,2,2,2,2,4,4,5,5,5,5,5) n=length(y) for (i in 1:n) #

On Sep 19, 2013, at 12:11 PM, Paul O'Rorke wrote: > So where does this leave things regards Eaton and shut down? Has anyone managed graceful shut down from an Eaton unit with either of the ushhid or bcmxcp drivers? Generally speaking, the UPS will signal when it reaches a low battery state, regardless of whether the UPS publishes runtime statistics, battery percentages, or any other

Hello, I'm trying to pull data from a network drive on a windows machine. The location is read into a string <rem> and then used later with a data input command. > rem<- "\\192.192.192.3\Shared\iris1.csv" Warning messages: 1: '\S' is an unrecognized escape in a character string 2: '\i' is an unrecognized escape in a character string 3: unrecognized

Hi, i'm having a serious problem with the following logon-script. It works under W2k, but not under W98, although W98 tells it is running startup-script .. First my smb.conf (only interesting parts) : ---snipp--- [global] server string = Samba-PDC %v on %h workgroup = WORKGROUP encrypt passwords = yes netbios name = 486dx66 domain master = Yes

Hello all: I am interested in using Speex for an application that streams audio from a (noisy) source, so I am interested in VAD and DTX operation. However, after browsing the archives of this list, I note that a number of people have not been satisfied with the operation of the VAD algorithm in Speex. This leads me to a few questions: - Is there a reference somewhere (other than the source

Hello all Nill experience with LVM. Have a Default Centos 4.4 install updated till a week ago with three HDDs. System's not booting up since my staff pulled out the plug due to a short circuit nearby. Machine is a PIII 550 MHz, with 3 HDDs 40 GB, 120 GB (Actually is bigger but my bios detects only upto 120 GB) & 20 GB...about half filled with data & the backup server taken out in

I'm trying to spec out hardware for a new office, and I'd like to include power over Ethernet as an option. I've seen a handful of PoE injectors around $1000 for 24 ports and a couple switches up around $2500 for 24 ports. Are there any cheaper options, short of buying a boatload of 1-port injectors off of ebay? I don't really need more then 24 ports of PoE out of 48 total

Hi Tom, Thanks for taking the time to answer this. On 11-01-16 18:10, Tom Stellard wrote: > On Mon, Jan 11, 2016 at 12:07:14PM +0100, Hans de Goede wrote: >> Hi, >> >> After a few distractions I'm back to work on the llvm tgsi backend. I've >> added clang integration and I can now compile a simple opencl program >> to something which sort of looks like