/* * netsniff-ng - the packet sniffing beast * Copyright 2011 - 2013 Daniel Borkmann. * Subject to the GPL, version 2. */ #define _GNU_SOURCE #include #include #include #include #include #include #include #include "pcap_io.h" #include "built_in.h" #include "ioops.h" #include "iosched.h" static size_t map_size = 0; static char *ptr_va_start, *ptr_va_curr; static void __pcap_mmap_write_need_remap(int fd) { int ret; off_t pos, map_size_old = map_size; off_t offset = ptr_va_curr - ptr_va_start; map_size = PAGE_ALIGN(map_size_old * 10 / 8); pos = lseek(fd, map_size, SEEK_SET); if (pos < 0) panic("Cannot lseek pcap file!\n"); ret = write_or_die(fd, "", 1); if (ret != 1) panic("Cannot write file!\n"); ptr_va_start = mremap(ptr_va_start, map_size_old, map_size, MREMAP_MAYMOVE); if (ptr_va_start == MAP_FAILED) panic("mmap of file failed!"); ret = madvise(ptr_va_start, map_size, MADV_SEQUENTIAL); if (ret < 0) panic("Failed to give kernel mmap advise!\n"); ptr_va_curr = ptr_va_start + offset; } static ssize_t pcap_mm_write(int fd, pcap_pkthdr_t *phdr, enum pcap_type type, const uint8_t *packet, size_t len) { size_t hdrsize = pcap_get_hdr_length(phdr, type); if ((off_t) (ptr_va_curr - ptr_va_start) + hdrsize + len > map_size) __pcap_mmap_write_need_remap(fd); fmemcpy(ptr_va_curr, &phdr->raw, hdrsize); ptr_va_curr += hdrsize; fmemcpy(ptr_va_curr, packet, len); ptr_va_curr += len; return hdrsize + len; } static ssize_t pcap_mm_read(int fd __maybe_unused, pcap_pkthdr_t *phdr, enum pcap_type type, uint8_t *packet, size_t len) { size_t hdrsize = pcap_get_hdr_length(phdr, type), hdrlen; if (unlikely((off_t) (ptr_va_curr + hdrsize - ptr_va_start) > map_size)) return -EIO; fmemcpy(&phdr->raw, ptr_va_curr, hdrsize); ptr_va_curr += hdrsize; hdrlen = pcap_get_length(phdr, type); if (unlikely((off_t) (ptr_va_curr + hdrlen - ptr_va_start) > map_size)) return -EIO; if (unlikely(hdrlen == 0 || hdrlen > len)) return -EINVAL; fmemcpy(packet, ptr_va_curr, hdrlen); ptr_va_curr += hdrlen; return hdrsize + hdrlen; } static inline off_t ____get_map_size(bool jumbo) { int allocsz = jumbo ? 16 : 3; return PAGE_ALIGN(sizeof(struct pcap_filehdr) + (PAGE_SIZE * allocsz) * 1024); } static void __pcap_mm_prepare_access_wr(int fd, bool jumbo) { int ret; off_t pos; struct stat sb; map_size = ____get_map_size(jumbo); ret = fstat(fd, &sb); if (ret < 0) panic("Cannot fstat pcap file!\n"); if (!S_ISREG (sb.st_mode)) panic("pcap dump file is not a regular file!\n"); pos = lseek(fd, map_size, SEEK_SET); if (pos < 0) panic("Cannot lseek pcap file!\n"); ret = write_or_die(fd, "", 1); if (ret != 1) panic("Cannot write file!\n"); ptr_va_start = mmap(NULL, map_size, PROT_WRITE, MAP_SHARED, fd, 0); if (ptr_va_start == MAP_FAILED) panic("mmap of file failed!"); ret = madvise(ptr_va_start, map_size, MADV_SEQUENTIAL); if (ret < 0) panic("Failed to give kernel mmap advise!\n"); ptr_va_curr = ptr_va_start + sizeof(struct pcap_filehdr); } static void __pcap_mm_prepare_access_rd(int fd) { int ret; struct stat sb; ret = fstat(fd, &sb); if (ret < 0) panic("Cannot fstat pcap file!\n"); if (!S_ISREG (sb.st_mode)) panic("pcap dump file is not a regular file!\n"); map_size = sb.st_size; ptr_va_start = mmap(NULL, map_size, PROT_READ, MAP_SHARED | MAP_LOCKED, fd, 0); if (ptr_va_start == MAP_FAILED) panic("mmap of file failed!"); ret = madvise(ptr_va_start, map_size, MADV_SEQUENTIAL); if (ret < 0) panic("Failed to give kernel mmap advise!\n"); ptr_va_curr = ptr_va_start + sizeof(struct pcap_filehdr); } static void pcap_mm_init_once(void) { set_ioprio_be(); } static int pcap_mm_prepare_access(int fd, enum pcap_mode mode, bool jumbo) { switch (mode) { case PCAP_MODE_RD: __pcap_mm_prepare_access_rd(fd); break; case PCAP_MODE_WR: __pcap_mm_prepare_access_wr(fd, jumbo); break; default: bug(); } return 0; } static void pcap_mm_fsync(int fd __maybe_unused) { msync(ptr_va_start, (off_t) (ptr_va_curr - ptr_va_start), MS_ASYNC); } static void pcap_mm_prepare_close(int fd, enum pcap_mode mode) { int ret; ret = munmap(ptr_va_start, map_size); if (ret < 0) panic("Cannot unmap the pcap file!\n"); if (mode == PCAP_MODE_WR) { ret = ftruncate(fd, (off_t) (ptr_va_curr - ptr_va_start)); if (ret) panic("Cannot truncate the pcap file!\n"); } } const struct pcap_file_ops pcap_mm_ops = { .init_once_pcap = pcap_mm_init_once, .pull_fhdr_pcap = pcap_generic_pull_fhdr, .push_fhdr_pcap = pcap_generic_push_fhdr, .prepare_access_pcap = pcap_mm_prepare_access, .prepare_close_pcap = pcap_mm_prepare_close, .read_pcap = pcap_mm_read, .write_pcap = pcap_mm_write, .fsync_pcap = pcap_mm_fsync, }; value='0' selected='selected'>unified
authorSrinivas Pandruvada <srinivas.pandruvada@linux.intel.com>2017-02-03 14:18:39 -0800
committerRafael J. Wysocki <rafael.j.wysocki@intel.com>2017-02-04 00:11:08 +0100
commit6e978b22efa1db9f6e71b24440b5f1d93e968ee3 (patch)
treec666f7a26b860674848949e39a610222b0723f89 /tools/testing/selftests/ntb
parent3c223c19aea85d3dda1416c187915f4a30b04b1f (diff)
cpufreq: intel_pstate: Disable energy efficiency optimization
Some Kabylake desktop processors may not reach max turbo when running in HWP mode, even if running under sustained 100% utilization. This occurs when the HWP.EPP (Energy Performance Preference) is set to "balance_power" (0x80) -- the default on most systems. It occurs because the platform BIOS may erroneously enable an energy-efficiency setting -- MSR_IA32_POWER_CTL BIT-EE, which is not recommended to be enabled on this SKU. On the failing systems, this BIOS issue was not discovered when the desktop motherboard was tested with Windows, because the BIOS also neglects to provide the ACPI/CPPC table, that Windows requires to enable HWP, and so Windows runs in legacy P-state mode, where this setting has no effect. Linux' intel_pstate driver does not require ACPI/CPPC to enable HWP, and so it runs in HWP mode, exposing this incorrect BIOS configuration. There are several ways to address this problem. First, Linux can also run in legacy P-state mode on this system. As intel_pstate is how Linux enables HWP, booting with "intel_pstate=disable" will run in acpi-cpufreq/ondemand legacy p-state mode. Or second, the "performance" governor can be used with intel_pstate, which will modify HWP.EPP to 0. Or third, starting in 4.10, the /sys/devices/system/cpu/cpufreq/policy*/energy_performance_preference attribute in can be updated from "balance_power" to "performance". Or fourth, apply this patch, which fixes the erroneous setting of MSR_IA32_POWER_CTL BIT_EE on this model, allowing the default configuration to function as designed. Signed-off-by: Srinivas Pandruvada <srinivas.pandruvada@linux.intel.com> Reviewed-by: Len Brown <len.brown@intel.com> Cc: 4.6+ <stable@vger.kernel.org> # 4.6+ Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Diffstat (limited to 'tools/testing/selftests/ntb')