Skip to content

Releases: rhboot/fwupdate

fwupdate 12

28 Jun 18:33
Compare
Choose a tag to compare
  • Better CI
  • Limit how large our WMI buffer can be
  • Add fwup_version() because fwupd wants it
  • Fix linking on (some) platforms that need libgcc_eh.a
  • abicheck improvements
  • Default to creating a UX capsule, whatever it may actually do.

fwupdate 11

09 Apr 18:47
11
Compare
Choose a tag to compare
  • lots of fixes from cov-scan and clang analyzer
  • support for Lenovo machines

fwupdate 10

08 Jan 22:27
10
Compare
Choose a tag to compare

fwupdate 10 - lots and lots of stuff

  • experimental support for UI Capsules
  • Dell WMI support
  • lots of bugfixes
  • configurable EFI ESP location by setting ESPMOUNTPOINT or the git
    config property fwupdate.espmountdir during the build.
  • Lots of coverity work
  • ABI compatibility checking during the release process
  • Make subdirectory builds work (see e24c4f8)

fwupdate 9

06 Mar 20:41
9
Compare
Choose a tag to compare
fwupdate 9 Pre-release
Pre-release

Update to fwupdate 9

  • gcc 7 fixes
  • Ensure BootNext entries are in BootOrder to work around some broken
    firmwares.
  • Minor bug fixes
  • Support for detecting that an admin password is set when trying to enable
    firmware features.

fwupdate 8

19 Aug 15:17
8
Compare
Choose a tag to compare
  • Fix some i686 build errors
  • Be less stupid about SONAMEs so in the future we'll only have to rebuild
    dependent things on actual ABI changes.
  • Only depend on libsmbios on x86, for now, because it hasn't been ported to
    Aarch64.

fwupdate 7

17 Aug 13:30
7
Compare
Choose a tag to compare
  • Fix the fix for ae7b85
  • fix one place where a second "rc" varibale is clobbering a result.

fwupdate 6

16 Aug 19:31
6
Compare
Choose a tag to compare
fwupdate 6 Pre-release
Pre-release
  • lots of build fixes for newer compilers and such
  • Use libsmbios on some systems to enable firmware updates (Mario Limonciello)
  • Use the correct reset type from the QueryCapsuleInfo data
  • Lots of fixes from auditing
  • Use efivar's error reporting infrastructure

Version 0.5!

18 Nov 15:56
0.5
Compare
Choose a tag to compare

Highlights in 0.5:

  • fwupdate.efi is called fwup$EFI_ARCH.efi now so weird platforms can
    have them coexist. "Platform" here might mean "distro tools that care
    about multilib". Anyway, it's needed to support things like baytrail.
  • Worked around shim command line bug where we need to treat
    LOAD_OPTIONS differently if we're invoked from the shell vs BDS
  • various debug features - SHIM_DEBUG and FWUPDATE_VERBOSE UEFI
    variables that'll let you get some debugging info some times
  • oh yeah, the actual debuginfo is useful
  • Automatically cleans up old instances on fresh OS installs
  • valgrind --leak-check=full on fwupdate doesn't show any errors at all
  • covscan shows only two things; one really doesn't matter, the other
    is because it doesn't understand our firmware variable data structure
    and can't work out that we have guaranteed the length of some data in
    a code path it isn't considering.
  • fwup_set_up_update() API improvements
  • killed fwup_sterror() and friends entirely
  • Should work on x64, ia32, and aarch64.

TODO for 1.0:

  • security audit for efi/fwupdate.c with secure boot violation in mind
  • debug Canoe Pass

fwupdate version 0.4!

02 Jun 18:42
0.4
Compare
Choose a tag to compare
  • Set DESTDIR so it's more consistently respected
  • Always use upper case for Boot#### names.
  • Create abbreviated device paths for our BootNext entry.
  • Make subdir Makefiles get the version right.
  • Fix ucs2len() to handle max=-1 correctly.
  • Compare the right blobs when we're searching old boot entries.
  • Fix .efi generation on non-x86 platforms.
  • Use a relative path for fwupdate.efi when launched from shim.
  • Show fewer debugging messages.
  • Set BootNext when we find an old Boot#### variable as well.
  • Add fwup_get_fw_type().

fwupdate 0.3 !

01 Jun 15:27
0.3
Compare
Choose a tag to compare
  • it seems to work on at least one system and one vm
  • it has some documentation
  • it's called 0.3 for a reason