x86info-1.31+git.20200121-bp156.1.3

- Add patch use-python3.patch:
  * Use python3 to create header files.

- Upgraded to new version 1.31+git.20200121
  * update intel codenames/models
  * gcc-8 fix
  * No more periodic versioned releases.
  [x86info-1.31+git.20200121.tar.xz]

- Use noun phrase in summary.

- Update to version 1.30+git.20180323:
  * rename release target to tarball
  * bump minimum for -Werror to gcc 6.3
  * add tag target
  * Make cpuid-freebsd.c:cpuid() handle top word of idx as count (%ecx), same as cpuid-linux.c:cpuid() does.
  * Make native_cpuid() handle top word of idx as count (%ecx), same as cpuid-linux.c:cpuid() does.
  * Fix vendor/inte/identify-family6-extended.c::get_namestring() to print cpu mode name correcty. When I tested on Kaby Lake machine, it shows:
  * amd: Identify Family 17h model CPUs
  * Makefile: Use pkg-config libpci location
- Use more accurate versioning
- Update build requirements
- Make building more verbose

- Update to HEAD (9501749231164135db25c72cbeb42bf32af1519b)
  * Make it possible to statically link with libpci.
  * Add another Nehalem variant
  * Add exact steppings and complete list of all bloomfield cores.
  * add Knights Mill Xeon Phi.
  * - Add PREFETCHW instruction flag.
  - Add Hardware Duty Cycling flag.
  - Add HW-controlled performance states (HWP) flags.
  * Add 0x64, 0xa0, 0xc3 and 0xc4 into DTLB_cache_table[].
  * - Add Atom E3500.
  - Add Atom X3-C3000 [Silvermont].
  * Add NetBSD support. Note that NetBSD has not API to access MSR register from userland.
  * handle null EBDA pointer.
  * handle failure from enumerate_cpus

- Update to 1.30
  * no upstream changelog available
- Cleanup spec file with spec-clener

- Use parallel build and remove ancient specfile tags and sections

- Update to latest git version
  - > 1.29 plus some more patches until
    git commit aa57556f60c89eb84f65cb55e0215ff9be97dbec
  Includes:
  - Show boost state info on AMD
  - Identify some more cpus
  - ...

- Update to version 1.28
  Changelog (extracted git commit subjects, shortened):
    Some more IDA decoding
    Start decoding IDA
    gather topology info per-cpu
    Split intel info into basic/extended.
    only display extended family/model if set
    print out the BIOS programmed string by default too.
    Also clarify which is x86info's guess.
    put back some of the mptable verbose/silent code.
    rename bluesmoke to machine_check
    Split MSR-PM into thermal and performance
    IA32_PERF_STATUS is model specific.
    Dump APIC registers
    Add and show info about virtual and physical address sizes
    Decode MTRRphysBase and MTRRphysMask
    use MB macro, like in the kernel
    update cache descriptors
    Decode MTRRcap and MTRRdefType registers
    Make makenodes script more robust
    Add an Atom variant
    Add an Intel codename.
    Add an atom ident.
    Add some Nehalem core codenames.
    Add Core i7 informational URL.
    Report whether longnops are supported

- Update to version 1.27
  Changelog (extracted git commit subjects from 1.25-1.27):
    only support topology parsing on intel for now
    Factor out the topology printing code into its own routine
    add more Intel CPUID flags
    Add informational URL for Pentium III Dothan.
    fix invalid usage of printf
    cpuid.c: Use native cpuid call silently instead of failing
    Remove the message about cpuid_up being inaccurate.
    Makefile: Override dependency output file with -MF
    Add additional Intel CPUID flags
    x86info: show microcode patch level for AMD CPUs
    x86info: identify newer AMD CPUs (e.g. Istanbul, Phenom II)
    Don't specify any ld flags until the linking occurred.
    Display register info when flag is unknown.
    Additional flags and a little cleanup
    Enhance descriptions for Intel feature flags.
    ecx flags update patch
    move get_cpu_info_basics() prior to get_feature_flags()
    Full RELRO support. Because there's no reason not to.
    improve the ht siblings text
    Handle non-sequential socket numbers correctly.
    Only calculate MHz on the first cpu, unless --all-cpus is passed.
    remove extraneous space
    discrimination between HT and dual-core.
    Store the number of siblings.
    delete core dumps with make clean
    clean up the output for single cpu case.
    Fix segfault by sizing the sockets array correctly.
    correctly clear the sockets struct.
    Make the output even more fancy.
    print the MHz in the summary. Also change the summary text a little.
    Clean up exit codes
    Print out a summary footer.
    Introduce a --debug switch, and hide some stuff behind it.
    Print a summary of the information from topology discovery.
    Make the new cpuid4 & cpuid_count functions take a cpu number
    integer instead of a cpudata struct so they match the other cpuid
    functions.
    Intel CPU topology reworking.
    PATCH: add support for new cpuid descriptors
    fix up release to use new git syntax

- Update to version 1.25
  * Add submitted results from core2-duo-E2180
  * Several compile fixes

- Update to version 1.24
  o Add Core i7 cache descriptors.
  o Check for a minimum version of glibc, as the affinity calls
  o Bump copyright
  o Atom series 200
  o Atom series 300
  o Atom N270 series
  o Add Atom.
  o Add Nehalem.
  o Ugh, decoding this lot is going to be a nightmare.
  o Add some more quad-core2's
  o Add Intel® Core™2 Extreme Processor QX9775
  o Add all core2 quad definitions from 318727.pdf
  o Warn on redundant declarations.  Remove some duplicates.
  o Add some extra warnings (missing declarations) and fix up a
    bunch
  o Fix up warnings. Add -Werror to catch new ones.

- update to version 1.23

- use RPM_OPT_FLAGS
- update to version 1.21 (see ChangeLog for full list of changes)

- update to version 1.20

- Update to version 1.10 (see ChangeLog for full list of changes)

- Update to version 1.7 (see ChangeLog for full list of changes)

- Update to version 1.6 (see ChangeLog for full list of changes)

- Initial package for SuSE Linux (Version 1.5)