| Age | Commit message (Collapse) | Author | 
|---|
|  | Signed-off-by: Burt P <pburt0@gmail.com> | 
|  | Starting in Debian 9, ldconfig requires root, and so it can't be
used to get the GLIBC version. Try `ldd --version` first. It should
give the same results.
Signed-off-by: Burt P <pburt0@gmail.com> | 
|  | Remove old notes.
Signed-off-by: Burt P <pburt0@gmail.com> | 
|  | Signed-off-by: Burt P <pburt0@gmail.com> | 
|  | Signed-off-by: Burt P <pburt0@gmail.com> | 
|  | * Of course, symbols and aliases are are the same
  as in /__symbols__ and /aliases. It still helps a bit.
Signed-off-by: Burt P <pburt0@gmail.com> | 
|  | * /aliases/* and /__symbols/* are always strings
* phandle ref element bug fix, re-enabled phandle refs
* notes on a few more property types
Signed-off-by: Burt P <pburt0@gmail.com> | 
|  | Signed-off-by: Burt P <pburt0@gmail.com> | 
|  | * dtr_str() requires obj, which knows its dtr.
Signed-off-by: Burt P <pburt0@gmail.com> | 
|  | * Moved device tree functions to
  modules/devices/devicetree/dt_util.c
* The dtr_* functions usable from outside devicetree.c,
  for example in get_motherboard().
  Must #include "dt_util.h"
* Now possible to use an alternate device tree root for testing
  -DOVRDTRROOT=\"/some/path\"
* Alternately, pass dtr_new() an alternate base path.
* Abandoned the tuple grouping and inherited properties stuff
  for now.
Signed-off-by: Burt P <pburt0@gmail.com> | 
|  | Signed-off-by: Burt P <pburt0@gmail.com> | 
|  | Signed-off-by: Burt P <pburt0@gmail.com> | 
|  | Signed-off-by: Burt P <pburt0@gmail.com> | 
|  | Signed-off-by: Burt P <pburt0@gmail.com> | 
|  | Based on information from
http://elinux.org/Device_Tree_Mysteries#.23xxx-cells_property_name
which is incomplete, try and group tuples using the relevant
inherited #(.*)-cells properies.
Just reg, clocks, and gpios so far. More possible.
See dt_tup_len().
Signed-off-by: Burt P <pburt0@gmail.com> | 
|  | Signed-off-by: Burt P <pburt0@gmail.com> | 
|  | Signed-off-by: Burt P <pburt0@gmail.com> | 
|  | Signed-off-by: Burt P <pburt0@gmail.com> | 
|  | Signed-off-by: Burt P <pburt0@gmail.com> | 
|  | Signed-off-by: Burt P <pburt0@gmail.com> | 
|  | The note added in
3824849f670ea9249a35a72b1f8f475e5f3fdea3
was because of a bug that was fixed in
1b66eb59292431d3786437e48424a1423a49a5d5
Signed-off-by: Burt P <pburt0@gmail.com> | 
|  | Signed-off-by: Burt P <pburt0@gmail.com> | 
|  | Signed-off-by: Burt P <pburt0@gmail.com> | 
|  | Signed-off-by: Burt P <pburt0@gmail.com> | 
|  | Signed-off-by: Burt P <pburt0@gmail.com> | 
|  | Initialize full_fmt to NULL so free() won't crash if it wasn't used.
Signed-off-by: Burt P <pburt0@gmail.com> | 
|  | Fixes a bug where a Raspberry Pi with a generic arm kernel will
be detected but not have a revision code in cpuinfo to lookup
information with.
Signed-off-by: Burt P <pburt0@gmail.com> | 
|  | Signed-off-by: Burt P <pburt0@gmail.com> | 
|  | The complete tree is now available to view. There are some bugs,
mostly to do with hardinfo protocol quirks.
Signed-off-by: Burt P <pburt0@gmail.com> | 
|  | Plus a couple other minor tweaks.
Signed-off-by: Burt P <pburt0@gmail.com> | 
|  | Signed-off-by: Burt P <pburt0@gmail.com> | 
|  | Hack a way to skip unavailable entries. Now "Resources" is
avaialbe outside of x86 when it wasn't before.
Signed-off-by: Burt P <pburt0@gmail.com> | 
|  | Signed-off-by: Burt P <pburt0@gmail.com> | 
|  | Signed-off-by: Burt P <pburt0@gmail.com> | 
|  | Signed-off-by: Burt P <pburt0@gmail.com> | 
|  | * Avoid using GHashTable
* Use single flag lookup table with bug: and pm: prefixes
  to avoid flag name collisions
Signed-off-by: Burt P <pburt0@gmail.com> | 
|  | Signed-off-by: Burt P <pburt0@gmail.com> | 
|  | Signed-off-by: Burt P <pburt0@gmail.com> | 
|  | * flag defintion strings are grouped into contexts:
  x86-flag, arm-flag, riscv-ext
* a comment is included for translators so that they know what flag
  the string defines without looking at the source code
Signed-off-by: Burt P <pburt0@gmail.com> | 
|  | Signed-off-by: Burt P <pburt0@gmail.com> | 
|  | Attempt adding support for pgettext without
the bulk of standard gettext.h
Signed-off-by: Burt P <pburt0@gmail.com> | 
|  | Signed-off-by: Burt P <pburt0@gmail.com> | 
|  | Note: "Computer" summary (not Computer->Summary).
Signed-off-by: Burt P <pburt0@gmail.com> | 
|  | * Python2 detection needs to look at stderr instead of stdout
* Python3 detection looks at stdout
* Perl6 detection
* scan_dev() now captures stdout and stderr and ignores the one
  it doesn't want. This prevents messages from appearing in the
  console when the scan happens.
Signed-off-by: Burt P <pburt0@gmail.com> | 
|  | Signed-off-by: Burt P <pburt0@gmail.com> | 
|  | Signed-off-by: Burt P <pburt0@gmail.com> | 
|  | Signed-off-by: Burt P <pburt0@gmail.com> | 
|  | An attempt to make updatepo.sh generate hardinfo.pot and .po files
that are the same regardless of the machine, so that diffs are less
noisy and more useful.
Signed-off-by: Burt P <pburt0@gmail.com> | 
|  | Some noise in the diff due to automatic trailing whitespace removal.
Signed-off-by: Burt P <pburt0@gmail.com> | 
|  | Depends on working pgettext().
Signed-off-by: Burt P <pburt0@gmail.com> |