Search Results

Search found 7213 results on 289 pages for 'multi processor'.

Page 184/289 | < Previous Page | 180 181 182 183 184 185 186 187 188 189 190 191  | Next Page >

• When compiling programs to run inside a VM, what should march and mtune be set to?

  - by Russ

  With VMs being slave to whatever the host machine is providing, what compiler flags should be provided to gcc? I would normally think that -march=native would be what you would use when compiling for a dedicated box, but the fine detail that -march=native is going to as indicated in this article makes me extremely wary of using it. So... what to set -march and -mtune to inside a VM? For a specific example... My specific case right now is compiling python (and more) in a linux guest inside a KVM-based "cloud" host that I have no real control over the host hardware (aside from 'simple' stuff like CPU GHz m CPU count, and available RAM). Currently, cpuinfo tells me I've got an "AMD Opteron(tm) Processor 6176" but I honestly don't know (yet) if that is reliable and whether the guest can get moved around to different architectures on me to meet the host's infrastructure shuffling needs (sounds hairy/unlikely). All I can really guarantee is my OS, which is a 64-bit linux kernel where uname -m yields x86_64.

  Read the article

• Why can't my networks reach each other?

  - by HOLOGRAPHICpizza

  We have two Buffalo WZR-HP-G300NH2 routers, with the default firmware, DD-WRT v24SP2-MULTI (10/31/11) std - build 17798. Each has a separate cable internet connection with a public static IP address. They are both in the 24.123.68.0/24 space. Both of them can contact pretty much the whole internet, and they can both be accessed out on the internet with no problem, but for some reason they can't talk to each other! When I try to ping one from the other I always get "Destination Host Unreachable". There are no strange routing or firewall rules in place. And they are both set to respond to pings, I can ping them from outside. Our main IT guy is going to call our ISP on Monday, but I'm impatient, so does anyone have any ideas?

  Read the article

• Need to detect the same application open on another computer on the network. Any software around tha

  - by Joe Schmoe

  I have a time management application that I use at home quite a lot and have running most of the time. At home, I have a desktop PC and a couple of laptops scattered around the house...all networked together. Unfortunately, the application I use is not multi-user and I risk losing/corrupting data if it has been left running on one computer inadvertently while I start using it on another one in another part of the house. I use Live Mesh to automatically keep the application's database synced across the different computers and I just need some way of making sure that I don't start using the application on another computer before closing it down on the previous one. Anyone know of any Windows software that can detect if an application is running simultaneously on different computers on my network, and warn me if I am about to have two open at the same time?

  Read the article

• How to troubleshoot if a zip file is valid or if it is big file size to be unzipped ?

  - by mireille raad

  Hello , I am trying to unzip a file with the size of 2GB I am getting the following error : unzip CLTE_C_08.zip Archive: CLTE_C_08.zip End-of-central-directory signature not found. Either this file is not a zipfile, or it constitutes one disk of a multi-part archive. In the latter case the central directory and zipfile comment will be found on the last disk(s) of this archive. unzip: cannot find zipfile directory in one of CLTE_C_08.zip or CLTE_C_08.zip.zip, and cannot find CLTE_C_08.zip.ZIP, period. After some googling, some people say that this error is because the file is too big, others say because file is corrupt, others say that it could be a not unix archive. So my question , how to find out if file is valid archive file on my Centos and what is the command/trick to uncompress big files ( if any ) Thanks in advance :)

  Read the article

• Windows undetectable after interrupted chkdsk

  - by Felthragar

  I have a computer that is running Windows XP. For some reason, the other day it wouldn't start giving me the following message: "ntoskrnl.exe is missing or corrupt" So I put the XP disc in the tray and fired up the repair console and ran the following: chkdsk /r It was on for about eight hours and it got to about 52% I believe. Then there was a power outage and the computer shut down (obviously). Today when I was booting it, it isn't even detecting there's an OS anymore. If I boot the computer with no cd in the tray it says: "Reboot and select a proper Boot device or Insert Boot Media in selected Boot device and press a key" If I run the repair console, or the xp installation program it isn't finding any OS installations. Any ideas on what to do next? Any help is appreciated. Thanks! Update: After turning boot-time diagnostics on, I got this message when booting without cd (instead of the previous one): "Couldn't open drive multi(0)disk(0)rdisk(0)partition(2)"

  Read the article

• Ways to divide the high/low byte from a 16bit address?

  - by Grissiom

  Hello, I'm developing a software on 8051 processor. A frequent job is to divide the high and low byte of a 16bit address. I want to see there are how many ways to achieve it. The ways I come up so far are: (say ptr is a 16bit pointer, and int is 16bit int) ADDH = (unsigned int) ptr >> 8; ADDL = (unsigned int) ptr & 0x00FF; and ADDH = ((unsigned char *)&ptr)[0]; ADDL = ((unsigned char *)&ptr)[1]; Does anyone have any other bright ideas? ;) And anyone can tell me which way is more efficient?

  Read the article

• I added some options to stop spam with Postfix, but now won't send email to remote domains

  - by willdanceforfun

  I had a working Postfix server, but added a few lines to my main.cf in a hope to block some common spam. Those lines I added were: smtpd_helo_required = yes smtpd_recipient_restrictions = reject_invalid_hostname, reject_unknown_recipient_domain, reject_unauth_pipelining, permit_mynetworks, permit_sasl_authenticated, reject_unauth_destination, reject_rbl_client multi.uribl.com, reject_rbl_client dsn.rfc-ignorant.org, reject_rbl_client dul.dnsbl.sorbs.net, reject_rbl_client list.dsbl.org, reject_rbl_client sbl-xbl.spamhaus.org, reject_rbl_client bl.spamcop.net, reject_rbl_client dnsbl.sorbs.net, reject_rbl_client cbl.abuseat.org, reject_rbl_client ix.dnsbl.manitu.net, reject_rbl_client combined.rbl.msrbl.net, reject_rbl_client rabl.nuclearelephant.com, permit It appears my postfix is now receiving normal emails fine, and blocking spam emails. But when I now try to use this server myself to send to a remote domain (an email not on my server) I get bounced, with maillog saying something like this: Nov 12 06:19:36 srv postfix/smtpd[11756]: NOQUEUE: reject: RCPT from unknown[xx.xx.x.xxx]: 450 4.1.2 <[email protected]>: Recipient address rejected: Domain not found; from=<[email protected]> to=<[email protected]> proto=ESMTP helo=<[192.168.1.100]> Is that saying 'domain not found' for gmail.com? Why is that recipient address rejected? An output of my postconf-n is: alias_database = hash:/etc/aliases alias_maps = hash:/etc/aliases broken_sasl_auth_clients = yes command_directory = /usr/sbin config_directory = /etc/postfix daemon_directory = /usr/libexec/postfix data_directory = /var/lib/postfix debug_peer_level = 2 html_directory = no inet_interfaces = all inet_protocols = all mail_owner = postfix mailbox_size_limit = 0 mailq_path = /usr/bin/mailq.postfix manpage_directory = /usr/share/man mydestination = $myhostname, localhost.$mydomain, localhost, $mydomain mydomain = primarydomain.net myhostname = mail.primarydomain.net myorigin = $myhostname newaliases_path = /usr/bin/newaliases.postfix queue_directory = /var/spool/postfix readme_directory = /usr/share/doc/postfix-2.6.6/README_FILES relay_domains = $mydestination, primarydomain.net, secondarydomain.org sample_directory = /usr/share/doc/postfix-2.6.6/samples sendmail_path = /usr/sbin/sendmail.postfix setgid_group = postdrop smtpd_client_restrictions = permit_sasl_authenticated smtpd_helo_required = yes smtpd_recipient_restrictions = reject_invalid_hostname, reject_unknown_recipient_domain, reject_unauth_pipelining, permit_mynetworks, permit_sasl_authenticated, reject_unauth_destination, reject_rbl_client multi.uribl.com, reject_rbl_client dsn.rfc-ignorant.org, reject_rbl_client dul.dnsbl.sorbs.net, reject_rbl_client list.dsbl.org, reject_rbl_client sbl-xbl.spamhaus.org, reject_rbl_client bl.spamcop.net, reject_rbl_client dnsbl.sorbs.net, reject_rbl_client cbl.abuseat.org, reject_rbl_client ix.dnsbl.manitu.net, reject_rbl_client combined.rbl.msrbl.net, reject_rbl_client rabl.nuclearelephant.com, permit smtpd_sasl_auth_enable = yes smtpd_sasl_path = private/auth smtpd_sasl_type = dovecot smtpd_sender_restrictions = reject_unknown_sender_domain soft_bounce = no unknown_local_recipient_reject_code = 550 virtual_alias_domains = mail.secondarydomain.org virtual_alias_maps = hash:/etc/postfix/virtual Any insight greatly appreciated. Edit: here is the dig mx gmail.com from the server: ; <<>> DiG 9.8.2rc1-RedHat-9.8.2-0.17.rc1.el6_4.4 <<>> mx gmail.com ;; global options: +cmd ;; Got answer: ;; ->>HEADER<<- opcode: QUERY, status: NOERROR, id: 31766 ;; flags: qr rd ra; QUERY: 1, ANSWER: 5, AUTHORITY: 4, ADDITIONAL: 14 ;; QUESTION SECTION: ;gmail.com. IN MX ;; ANSWER SECTION: gmail.com. 1207 IN MX 5 gmail-smtp-in.l.google.com. gmail.com. 1207 IN MX 30 alt3.gmail-smtp-in.l.google.com. gmail.com. 1207 IN MX 20 alt2.gmail-smtp-in.l.google.com. gmail.com. 1207 IN MX 40 alt4.gmail-smtp-in.l.google.com. gmail.com. 1207 IN MX 10 alt1.gmail-smtp-in.l.google.com. ;; AUTHORITY SECTION: gmail.com. 109168 IN NS ns1.google.com. gmail.com. 109168 IN NS ns4.google.com. gmail.com. 109168 IN NS ns3.google.com. gmail.com. 109168 IN NS ns2.google.com. ;; ADDITIONAL SECTION: alt1.gmail-smtp-in.l.google.com. 207 IN A 173.194.70.27 alt1.gmail-smtp-in.l.google.com. 248 IN AAAA 2a00:1450:4001:c02::1b gmail-smtp-in.l.google.com. 200 IN A 173.194.67.26 gmail-smtp-in.l.google.com. 248 IN AAAA 2a00:1450:400c:c05::1b alt3.gmail-smtp-in.l.google.com. 207 IN A 74.125.143.27 alt3.gmail-smtp-in.l.google.com. 249 IN AAAA 2a00:1450:400c:c05::1b alt2.gmail-smtp-in.l.google.com. 207 IN A 173.194.69.27 alt2.gmail-smtp-in.l.google.com. 248 IN AAAA 2a00:1450:4008:c01::1b alt4.gmail-smtp-in.l.google.com. 207 IN A 173.194.79.27 alt4.gmail-smtp-in.l.google.com. 249 IN AAAA 2607:f8b0:400e:c01::1a ns2.google.com. 281970 IN A 216.239.34.10 ns3.google.com. 281970 IN A 216.239.36.10 ns4.google.com. 281970 IN A 216.239.38.10 ns1.google.com. 281970 IN A 216.239.32.10

  Read the article

• Bash - Test for multiple users

  - by Mike Purcell

  I am trying to test if current user one of two allowed users to start a process, but I can't seem to get the multi-condition to work correctly: test ($(whoami) != 'mpurcell' || $(whoami) != 'root')) && (echo "Cannot start script as non-ccast user..."; exit 1) Is there a way to test multiple users without have to enter two lines, like this: test $(whoami) != 'mpurcell' && (echo "Cannot start script as non-ccast user..."; exit 1) test $(whoami) != 'root' && (echo "Cannot start script as non-ccast user..."; exit 1)

  Read the article

• RandR 1.4 Optimus Dual Monitor

  - by mathepic

  So, I have a dual monitor setup. HDMI comes in through Nvidia, main display is through Intel (I think). I want to use XMonad with the dual setup, and I want to be able to run with or without the second monitor. Is this even doable? I'm using RandR 1.4 and can get both monitors to display something at the same time (by messing with xrandr) but XMonad can never detect more than one rectangle form Xinerama. Does anyone have a working multi-monitor xinerama or twinview configuration that works with optimus/randr 1.4?

  Read the article

• Increasing the JVM maximum heap size for memory intensive applications

  - by Alceu Costa

  I need to run a Java memory intensive application that uses more than 2GB, but I am having problems to increase the heap maximum size. So far, I have tried the following approaches: Setting the -Xmx parameter, e.g. -Xmx3000m. This approaches fails at the creation of the JVM. From what I've googled, it looks like that -Xmx must be less than 2GB. Using the -XX:+AggressiveHeap option. When I try this approach I get an 'Not enough memory' error that tells that the heap size is 1273.4 MB, even though my computer has 8GB of memory. Is there another approach that I can try to increase the maximum heap size of the JVM? Here's a summary of the computer specs: OS: Windows 7 (64 bit) Processor: Intel Core i7 (2.66 GHz) Memory: 8 GB

  Read the article

• StructureMap - Injecting a dependency into a base class?

  - by David

  In my domain I have a handful of "processor" classes which hold the bulk of the business logic. Using StructureMap with default conventions, I inject repositories into those classes for their various IO (databases, file system, etc.). For example: public interface IHelloWorldProcessor { string HelloWorld(); } public class HelloWorldProcessor : IHelloWorldProcessor { private IDBRepository _dbRepository; public HelloWorldProcessor(IDBRepository dbRepository) { _dbRepository = dbrepository; } public string HelloWorld(){ return _dbRepository.GetHelloWorld(); } } Now, there are some repositories that I'd like to be available to all processors, so I made a base class like this: public class BaseProcessor { protected ICommonRepository _commonRepository; public BaseProcessor(ICommonRepository commonRepository) { _commonRepository = commonRepository; } } But when my other processors inherit from it, I get a compiler error on each one saying that there's no constructor for BaseProcessor which takes zero arguments. Is there a way to do what I'm trying to do here? That is, to have common dependencies injected into a base class that my other classes can use without having to write the injections into each one?

  Read the article

• Configuration tools for multiple monitors for X / Linux

  - by richard

  I have Ubuntu 10.04 running gnome and two monitors. I am wondering if a can get a better multi-monitor configuration tool. The one I have, gnome-display-properties, has too many problems, including: when I swapped my monitors over, the narrower one now on the left. There is a width calculation error, such that I have a virtual monitor the width of the wide-monitor on the narrow-monitor and part of the wide monitor. And a virtual narrow-monitor on the remainder of the wide-monitor. I would like: nobugs. to be able to select which is primary monitor. to have multiple configurations. configurations to be automatically selected based on which monitors are attached. configurations to be cycled (reliably) when display mode key is pressed. when a display is deactivated, for windows to migrate to remaining monitors. option to not change display resolution when mirroring, but to use side/top blanking bars to pad out screen.

  Read the article

• rsyslog forward all except ldap

  - by Brian

  I have Centos 6 servers running openLDAP. In the rsyslog.conf, I forward the logs to my central server with this line: *.* @10.10.10.10:514 openldap seems incredibly chatty. I have 3 servers in a multi-master cluster. Those 3 servers generate twice as many logs as my other 80 servers combined. I have been unsuccessful in figuring out how to tell openLDAP to use a sensible log level. (we never specifically set the log level) Since these are my main authentication sources, I'm a bit hesitant to "play around" with them. Is there a way to tell rsyslog to forward everything EXCEPT LOCAL4?

  Read the article

• Supporting Piping (A Useful Hello World)

  - by blastthisinferno

  I am trying to write a collection of simple C++ programs that follow the basic Unix philosophy by: Make each program do one thing well. Expect the output of every program to become the input to another, as yet unknown, program. I'm having an issue trying to get the output of one to be the input of the other, and getting the output of one be the input of a separate instance of itself. Very briefly, I have a program add which takes arguments and spits out the summation. I want to be able to pipe the output to another add instance. ./add 1 2 | ./add 3 4 That should yield 6 but currently yields 10. I've encountered two problems: The cin waits for user input from the console. I don't want this, and haven't been able to find a simple example showing a the use of standard input stream without querying the user in the console. If someone knows of an example please let me know. I can't figure out how to use standard input while supporting piping. Currently, it appears it does not work. If I issue the command ./add 1 2 | ./add 3 4 it results in 7. The relevant code is below: add.cpp snippet // ... COMMAND LINE PROCESSING ... std::vector<double> numbers = multi.getValue(); // using TCLAP for command line parsing if (numbers.size() > 0) { double sum = numbers[0]; double arg; for (int i=1; i < numbers.size(); i++) { arg = numbers[i]; sum += arg; } std::cout << sum << std::endl; } else { double input; // right now this is test code while I try and get standard input streaming working as expected while (std::cin) { std::cin >> input; std::cout << input << std::endl; } } // ... MORE IRRELEVANT CODE ... So, I guess my question(s) is does anyone see what is incorrect with this code in order to support piping standard input? Are there some well known (or hidden) resources that explain clearly how to implement an example application supporting the basic Unix philosophy? @Chris Lutz I've changed the code to what's below. The problem where cin still waits for user input on the console, and doesn't just take from the standard input passed from the pipe. Am I missing something trivial for handling this? I haven't tried Greg Hewgill's answer yet, but don't see how that would help since the issue is still with cin. // ... COMMAND LINE PROCESSING ... std::vector<double> numbers = multi.getValue(); // using TCLAP for command line parsing double sum = numbers[0]; double arg; for (int i=1; i < numbers.size(); i++) { arg = numbers[i]; sum += arg; } // right now this is test code while I try and get standard input streaming working as expected while (std::cin) { std::cin >> arg; std::cout << arg << std::endl; } std::cout << sum << std::endl; // ... MORE IRRELEVANT CODE ...

  Read the article

• Wordpress 3 mutli site install

  - by mike

  Hello, Trying to figure out if this is possible... My company has a cms product that was written in Java and we decided to use Wordpress to run blogs for our clients. Obviously, Wordpress does not run on tomcat(at least not by default) so we installed Pound(http://www.apsis.ch/pound/) on our server and have setup any Apache and Tomcat on different ports. When "/blog/" is requested, the request is directed to Apache. This works fine but we would like to use Wordpress multi site so that we can manage all the blogs from a single interface. We would also like the url for every site to be "/blog/" example: http://www.site1.com/blog/ http://www.site2.com/blog/ I'm thinking it would have to be done with apache??? Is it even possible? Thanks!

  Read the article

• How do I read multiple lines from STDIN into a variable?

  - by The Wicked Flea

  I've been googling this question to no avail. I'm automating a build process here at work, and all I'm trying to do is get version numbers and a tiny description of the build which may be multi-line. The system this runs on is OSX 10.6.8. I've seen everything from using CAT to processing each line as necessary. I can't figure out what I should use and why. Attempts read -d '' versionNotes Results in garbled input if the user has to use the backspace key. Also there's no good way to terminate the input as ^D doesn't terminate and ^C just exits the process. read -d 'END' versionNotes Works... but still garbles the input if the backspace key is needed. while read versionNotes do echo " $versionNotes" >> "source/application.yml" done Doesn't properly end the input (because I'm too late to look up matching against an empty string).

  Read the article

• Custom annotations to configure tests

  - by ace

  First of al let me start off by saying I think custom annotations can be used for this but i'm not totally sure. I would like to have a set of annotations that I can decorate some test classes with. The annotations would allow me to configure the test for different environments. Example: public class Atest extends BaseTest{ private String env; @Login(environment=env) public void testLogin(){ //do something } @SignUp(environment=env) public void testSignUp(){ //do something } } The idea here would be that the login annotation would then be used to lookup the username and password to be used in the testLogin method for testing a login process for a particular environment. So my question(s) is this possible to do with annotations? If so I have not been able to find a decent howto online to do something like this. Everything out there seems to be your basic here's how to do your custom annotations and a basic processor but I haven't found anything for a situation like this. Ideas?

  Read the article

• Looking for a central image database and tagging system for a group of users

  - by jstarek

  I'm doing IT support for a small volunteer organisation who needs to centrally store and organize around 2500 photos. Can anyone here recommend a database or similar system which matches the following criteria: Intuitive to use for users with little computer experience Multi-user support, ideally with integration in our existing LDAP user directory Should have a web-interface Not a hosted solution like Picasa (because we have a rather slow internet connection with very slow upload) Should allow tagging of images, sorting by various criteria and storing copyright information If there are native GNOME and/or Windows clients for the tool, that would be a great benefit. Many thanks in advance!

  Read the article

• Use ImageMagick to convert TIFF to PNGs, how to improve the speed?

  - by Woo

  I am using "convert" from IM to get PNGs from multi-page TIFF files, everything is good except the speed. From "convert" documentation, I found: For the MNG and PNG image formats, the quality value sets the zlib compression level (quality / 10) and filter-type (quality % 10). For compression level 0, the Huffman-only strategy is used, which is fastest but not necessarily the worst compression. The default PNG compression is 75. So I tried "-quality 0", but almost no changes with the spreed. Anyone can share the ideas of how to improve the spreed? Here are my command: convert 100Pages.tif[0,1,2,3,4,5] -quality 0 100Pages.png Thanks!

  Read the article

• Parallel Haskell in order to find the divisors of a huge number

  - by Dragno

  I have written the following program using Parallel Haskell to find the divisors of 1 billion. import Control.Parallel parfindDivisors :: Integer->[Integer] parfindDivisors n = f1 `par` (f2 `par` (f1 ++ f2)) where f1=filter g [1..(quot n 4)] f2=filter g [(quot n 4)+1..(quot n 2)] g z = n `rem` z == 0 main = print (parfindDivisors 1000000000) I've compiled the program with ghc -rtsopts -threaded findDivisors.hs and I run it with: findDivisors.exe +RTS -s -N2 -RTS I have found a 50% speedup compared to the simple version which is this: findDivisors :: Integer->[Integer] findDivisors n = filter g [1..(quot n 2)] where g z = n `rem` z == 0 My processor is a dual core 2 duo from Intel. I was wondering if there can be any improvement in above code. Because in the statistics that program prints says: Parallel GC work balance: 1.01 (16940708 / 16772868, ideal 2) and SPARKS: 2 (1 converted, 0 overflowed, 0 dud, 0 GC'd, 1 fizzled) What are these converted , overflowed , dud, GC'd, fizzled and how can help to improve the time.

  Read the article

• Cheapest High Available Web Server [closed]

  - by xyz

  I would like to create a high-available setup (e.g. a small cluster) for a webserver, i.e. it will run Apache, PHP and MySQL. There will be between 2-8 small websites running with only very little traffic and workload. High availability is however very important. I don't want to be dependent on 1 datacenter, so there must be a minimum of 2 servers placed in different datacenters, and if one server goes down, the user must experience no or only a minimum of downtime - and no data loss. I have considered Amazon AWS using their Elastic Load Balancing, since it is possible to buy 2 EC2 instances in 2 availability zones and set up load balancing and RDS (Multi-AZ). However this seems rather expensive. Using the AWS price calculator http://calculator.s3.amazonaws.com/calc5.html it totals to 185$/month the first year (including the free tier). Are my calculations incorrect or is there a cheaper way to make this HA setup? Best regards

  Read the article

• what is value of x for load and store

  - by Kevinniceguy

  This is some challenge On a single processor system, in which load and store are assumed to be atomic, what are all the possible values for x after both threads have completed in the following execution, assuming that x is initialised to O? Hint: you need to consider how this code might be compiled into machine language. for (int i = 0; i < 5; i++) : x = x + 1; for (int j = 0; j < 5; j++) : x = x + 1;

  Read the article

• Can I use the snipping tool to take a screenshot of the windows 8 start screen or modern apps?

  - by Journeyman Geek

  One of the tools I've found invaluable in answering questions on SU is the snipping tool. I may on occasion need to take screenshots of part of the start screen or 'modern' apps. I may not want to take a complete screenshot, and while I can use PrtSc and switch back into desktop to paste it, this is clunky if I need to document a multi-step process. Can I use the snipping tool on modern apps or the start screen? If not is there a configurable way to save a series of screenshots to a fixed folder, say when I press a combination of keys, so I can work, screenshot, then crop and annotate the folder of images?

  Read the article

• Odd optimization problem under MSVC

  - by Goz

  I've seen this blog: http://igoro.com/archive/gallery-of-processor-cache-effects/ The "weirdness" in part 7 is what caught my interest. My first thought was "Thats just C# being weird". Its not I wrote the following C++ code. volatile int* p = (volatile int*)_aligned_malloc( sizeof( int ) * 8, 64 ); memset( (void*)p, 0, sizeof( int ) * 8 ); double dStart = t.GetTime(); for (int i = 0; i < 200000000; i++) { //p[0]++;p[1]++;p[2]++;p[3]++; // Option 1 //p[0]++;p[2]++;p[4]++;p[6]++; // Option 2 p[0]++;p[2]++; // Option 3 } double dTime = t.GetTime() - dStart; The timing I get on my 2.4 Ghz Core 2 Quad go as follows: Option 1 = ~8 cycles per loop. Option 2 = ~4 cycles per loop. Option 3 = ~6 cycles per loop. Now This is confusing. My reasoning behind the difference comes down to the cache write latency (3 cycles) on my chip and an assumption that the cache has a 128-bit write port (This is pure guess work on my part). On that basis in Option 1: It will increment p[0] (1 cycle) then increment p[2] (1 cycle) then it has to wait 1 cycle (for cache) then p[1] (1 cycle) then wait 1 cycle (for cache) then p[3] (1 cycle). Finally 2 cycles for increment and jump (Though its usually implemented as decrement and jump). This gives a total of 8 cycles. In Option 2: It can increment p[0] and p[4] in one cycle then increment p[2] and p[6] in another cycle. Then 2 cycles for subtract and jump. No waits needed on cache. Total 4 cycles. In option 3: It can increment p[0] then has to wait 2 cycles then increment p[2] then subtract and jump. The problem is if you set case 3 to increment p[0] and p[4] it STILL takes 6 cycles (which kinda blows my 128-bit read/write port out of the water). So ... can anyone tell me what the hell is going on here? Why DOES case 3 take longer? Also I'd love to know what I've got wrong in my thinking above, as i obviously have something wrong! Any ideas would be much appreciated! :) It'd also be interesting to see how GCC or any other compiler copes with it as well! Edit: Jerry Coffin's idea gave me some thoughts. I've done some more tests (on a different machine so forgive the change in timings) with and without nops and with different counts of nops case 2 - 0.46 00401ABD jne (401AB0h) 0 nops - 0.68 00401AB7 jne (401AB0h) 1 nop - 0.61 00401AB8 jne (401AB0h) 2 nops - 0.636 00401AB9 jne (401AB0h) 3 nops - 0.632 00401ABA jne (401AB0h) 4 nops - 0.66 00401ABB jne (401AB0h) 5 nops - 0.52 00401ABC jne (401AB0h) 6 nops - 0.46 00401ABD jne (401AB0h) 7 nops - 0.46 00401ABE jne (401AB0h) 8 nops - 0.46 00401ABF jne (401AB0h) 9 nops - 0.55 00401AC0 jne (401AB0h) I've included the jump statetements so you can see that the source and destination are in one cache line. You can also see that we start to get a difference when we are 13 bytes or more apart. Until we hit 16 ... then it all goes wrong. So Jerry isn't right (though his suggestion DOES help a bit), however something IS going on. I'm more and more intrigued to try and figure out what it is now. It does appear to be more some sort of memory alignment oddity rather than some sort of instruction throughput oddity. Anyone want to explain this for an inquisitive mind? :D Edit 3: Interjay has a point on the unrolling that blows the previous edit out of the water. With an unrolled loop the performance does not improve. You need to add a nop in to make the gap between jump source and destination the same as for my good nop count above. Performance still sucks. Its interesting that I need 6 nops to improve performance though. I wonder how many nops the processor can issue per cycle? If its 3 then that account for the cache write latency ... But, if thats it, why is the latency occurring? Curiouser and curiouser ...

  Read the article

• Need help for this syntax: "#define LEDs (char *) 0x0003010"

  - by Noge

  I'm doing programming of a softcore processor, Nios II from Altera, below is the code in one of the tutorial, I manage to get the code working by testing it on the hardware (DE2 board), however, I could not understand the code. #define Switches (volatile char *) 0x0003000 #define LEDs (char *) 0x0003010 void main() { while (1) *LEDs = *Switches; } What I know about #define is that, it is either used to define a constant, or a macro, but why in the above code, there are casting like, (char *) 0x0003010, in #define? why the 2 constants, Switches and LEDs act like a variable instead of a constant? Thanks in advance !

  Read the article

< Previous Page | 180 181 182 183 184 185 186 187 188 189 190 191  | Next Page >