search for: number_of_cor

...ro cost. I would have a global number of shards that changes > rarely, and re-compute it on entry to each function with something along the > lines of: > > if (thread-ID != main's thread-ID && shard_count == 1) { >   shard_count = std::min(MAX, std::max(NUMBER_OF_THREADS, NUMBER_OF_CORES)); >   // if shard_count changed with this, we can also call a library routine here > that does the work of allocating the actual extra shards. > } Is it possible to hook on something more clever than function entry? Choosing to do this based on thread creation or something like that co...

...andler Carruth <chandlerc at google.com> wrote: > On Thu, Apr 17, 2014 at 1:27 PM, Justin Bogner <mail at justinbogner.com> wrote: > Chandler Carruth <chandlerc at google.com> writes: > > if (thread-ID != main's thread-ID && shard_count < std::min(MAX, NUMBER_OF_CORES)) { > > shard_count = std::min(MAX, std::max(NUMBER_OF_THREADS, NUMBER_OF_CORES)); > > // if shard_count changed with this, we can also call a library routine here > > that does the work of allocating the actual extra shards. > > } > > Note, I had a bug here. Th...

Hi, The current design of -fprofile-instr-generate has the same fundamental flaw as the old gcc's gcov instrumentation: it has contention on counters. A trivial synthetic test case was described here: http://lists.cs.uiuc.edu/pipermail/llvmdev/2013-October/066116.html For the problem to appear we need to have a hot function that is simultaneously executed by multiple threads -- then we will

...ocessor"); + if(!udev_enumerate_scan_devices(enumerator)) + { + struct udev_list_entry* entries = udev_enumerate_get_list_entry(enumerator); + struct udev_list_entry* entry; + + udev_list_entry_foreach(entry, entries) + { + core_count++; + } + set_number_of_cores(core_count); + } + + udev_enumerate_unref(enumerator); + udev_unref(udev); + + ifstream input("/proc/cpuinfo"); + if(input.is_open()) + { + string regexstr = "(.*\\S)\\s*:\\s*(\\S.*)"; + int expected = 3; + int found[expected * 3]; + const char* p...

...int core_count = 0; + int cpu_count = 0; string model = "unknown"; struct udev* udev = udev_new(); @@ -50,9 +50,9 @@ LinuxPlatform::LinuxPlatform() udev_list_entry_foreach(entry, entries) { - core_count++; + cpu_count++; } - set_number_of_cores(core_count); + setNumberOfCPUCores(cpu_count); } udev_enumerate_unref(enumerator); @@ -100,7 +100,7 @@ LinuxPlatform::LinuxPlatform() } else { - if(name == "model name") set_processor_model(value); +...

...systems (and amount of RAM!) is > O(NCORES), not O(1). In some read-mostly cases it's possible to achieve > O(1) memory consumption, and that's great. But if it's not the case here, > let it be so. > > > > > shard_count = std::min(MAX, std::max(NUMBER_OF_THREADS, NUMBER_OF_CORES)) > > Threads do not produce contention, it's cores that produce contention. > The formula must be: shard_count = k*NCORES > And if you want less memory in single-threaded case, then: shard_count = > min(k*NCORES, c*NTHREADS) > > > > >We are talking about develo...

...st. I would have a global number of shards that changes > rarely, and re-compute it on entry to each function with something along > the lines of: > > if (thread-ID != main's thread-ID && shard_count == 1) { > shard_count = std::min(MAX, std::max(NUMBER_OF_THREADS, > NUMBER_OF_CORES)); > // if shard_count changed with this, we can also call a library routine > here that does the work of allocating the actual extra shards. > } > > MAX is a fixed cap so even on systems with 100s of cores we don't do > something silly. NUBER_OF_THREADS, if supported on t...

...some read-mostly cases it's possible to achieve >>> O(1) memory consumption, and that's great. But if it's not the case here, >>> let it be so. >>> >>> >>> >>> > shard_count = std::min(MAX, std::max(NUMBER_OF_THREADS, >>> NUMBER_OF_CORES)) >>> >>> Threads do not produce contention, it's cores that produce contention. >>> The formula must be: shard_count = k*NCORES >>> And if you want less memory in single-threaded case, then: shard_count = >>> min(k*NCORES, c*NTHREADS) >>&gt...

...flags = (cpu.flags.length > 255) ? "#{cpu.flags[0..251]}..." : cpu.flags - detail = Cpu.new( - 'cpu_number' => cpu.cpunum, - 'core_number' => cpu.corenum, - 'number_of_cores' => cpu.numcores, - 'vendor' => cpu.vendor, - 'model' => cpu.model.to_s, - 'family' => cpu.family.to_s, - 'cpuid_level' =...

Removes node identification functionality from host-browser.rb and adds a new script, host-register.rb, that takes over that functionality. The chief difference is that host-browser used a simple TCP server setup to get data from the node, while host-register uses the qpid bus to do so. Specifically, it communicates with the matahari qmf agent added to the node in two related patchsets to node

I have tried 4 time unsuccessfully to install Xen 4 on Ubuntu Lucid (10.4) 64 bit version using the tutorial at http://bderzhavets.wordpress.com/2010/04/24/set-up-ubuntu-10-04-server-pv-domu-at-xen-4-0-dom0-pvops-2-6-32-10-kernel-dom0-on-top-of-ubuntu-10-04-server/ on a Lenovo thinkcentre workstation. When I boot I get a kernel panic. If I boot the Linux kernel that I compiled using the

Hello, I have been working on VGA-passthrough in the last month. When Xen was at his version 3.5 there was a few patches to make this work for some graphic cards using the patches developed by Han Weidong and posted here (http://lists.xensource.com/archives/html/xen-devel/2009-08/msg01176.html) and the same patches posted file by file are here

Hello, I have been working on VGA-passthrough in the last month. When Xen was at his version 3.5 there was a few patches to make this work for some graphic cards using the patches developed by Han Weidong and posted here (http://lists.xensource.com/archives/html/xen-devel/2009-08/msg01176.html) and the same patches posted file by file are here