On 14 March 2017 at 16:18, Paolo Bonzini <pbonz...@redhat.com> wrote: > From: Jitendra Kolhe <jitendra.ko...@hpe.com> > > Using "-mem-prealloc" option for a large guest leads to higher guest > start-up and migration time. This is because with "-mem-prealloc" option > qemu tries to map every guest page (create address translations), and > make sure the pages are available during runtime. virsh/libvirt by > default, seems to use "-mem-prealloc" option in case the guest is > configured to use huge pages. The patch tries to map all guest pages > simultaneously by spawning multiple threads. Currently limiting the > change to QEMU library functions on POSIX compliant host only, as we are > not sure if the problem exists on win32. Below are some stats with > "-mem-prealloc" option for guest configured to use huge pages.
> --- a/util/oslib-posix.c > +++ b/util/oslib-posix.c > @@ -55,6 +55,21 @@ > #include "qemu/error-report.h" > #endif > > +#define MAX_MEM_PREALLOC_THREAD_COUNT (MIN(sysconf(_SC_NPROCESSORS_ONLN), > 16)) sysconf() can fail, in which case it will return -1... > +static bool touch_all_pages(char *area, size_t hpagesize, size_t numpages, > + int smp_cpus) > +{ > + uint64_t numpages_per_thread, size_per_thread; > + char *addr = area; > + int i = 0; > + > + memset_thread_failed = false; > + memset_num_threads = MIN(smp_cpus, MAX_MEM_PREALLOC_THREAD_COUNT); ...and memset_num_threads (being signed) will also end up -1 here... > + memset_thread = g_new0(MemsetThread, memset_num_threads); ...which we then pass to g_new0(), which will not have good results. (Spotted by Coverity, CID 1372465.) Hiding a sysconf() behind a macro that looks like it's going to be a constant integer is a bit confusing, incidentally, and we only use it in one place. I'd just have a memset_num_threads() function which calls sysconf and determines the right number from that and smp_cpus and 16, and handles sysconf failing gracefully. thanks -- PMM