rockbox/apps/plugins, branch puzzles

[WIP] Port of Simon Tatham's Puzzle Collection

2016-11-24T21:23:09+00:00

Original revision: 5123b1bf68777ffa86e651f178046b26a87cf2d9

MIT Licensed. Some games still crash and others are unplayable due to
issues with controls. Still need a "real" polygon filling algorithm.

The following games are at least partially broken for various reasons:

Cube:     crash with certain settings
Galaxies: crash
Inertia:  crash
Keen:     input issues
Loopy:    weird stuff happens
Map:      crash on input
Mines:    weird stuff happens on target
Palisade: input issues
Signpost: crash on input
Solo:     input issues
Towers:   input and drawing issues
Train Tracks: drawing issues
Twiddle:  weird animation on target
Undead:   input and drawing issues
Unequal:  input and drawing issues
Untangle: input issues

All in all, about 40% of the games are at least partially broken.

Change-Id: I7c69b6860ab115f973c8d76799502e9bb3d52368

Add circle drawing/filling to xlcd

2016-11-20T20:13:13+00:00

Change-Id: I5d28ade42145d9d82babcf62c0db7948927cafec

XWorld: some fixes

2016-11-19T18:17:14+00:00

Fixes sound on most platforms, original root cause was bad menu code
as well as DMA callbacks taking too long. Worked around with smaller
chunk sizes. Permanent fix would include moving mixing out of the
callback. Rewrites input with code from rockboy/doom. Cherry-picks a
change from Gregory Montoir's `rawgl' to patch the code wheel
screen. Finally, adds a motion blur filter on select targets.

Change-Id: I8df549c923c5075800c6625c36c8202e53de1d27

pdbox: fix bug shown by new gcc warning

2016-10-12T14:52:49+00:00

Change-Id: Ife308a3090f9bc62f2e98d2d979fdf393c22054d

Rockboy: revert FS#11745 patch

2016-10-03T16:16:47+00:00

William Shipley sent this patch via mailing list. Original Message:
  "There was a patch to rockboy in 2010 that fixed a problem where
  the bottom tile row would be rendered at the top.

  Here's the relevant issue from that time:
  http://www.rockbox.org/tracker/11745

  Note that Stummi openly admits it was a hack, not a true fix.

  Sometime in the 6 years since, the root cause was fixed. Currently,
  rockboy on the Clip+ and other 64px models shows the top of the gb
  screen along the bottom of the display.

  I confirmed that reversing this patch fixes the issue on my own
  Clip+"

I don't have any of the mentioned targets, so someone else has to test.

Change-Id: Id29b78c4ece0af53ba1280029c18b76808a21da2

Fix misleading indentation as pointed out by the new gcc warning

2016-08-12T13:33:28+00:00

Change-Id: Ia9c0b09ffb57dccbc7d13fce3b3a8a7d46b84857

test_mem: print clock speed in MHz so it fits on smaller screens.

2016-08-03T18:26:54+00:00

Change-Id: I81383c115bd9abac44ffc2ecdd8807855c80cf0e

Fix endianness handling for sha1 (used by the otp plugin)

2016-06-28T11:33:50+00:00

Change-Id: I408d88d984f9026539c2d1bc2a9608243802c152

One-Time Password client (HOTP and TOTP)

2016-06-05T18:25:09+00:00

* Implements RFC 4226 (HOTP) and RFC 6238 (TOTP)

* Adds sha1.c to apps/plugins/lib (orignally tools/hmac-sha1.c)

* See manual entry for more information

Change-Id: Ia4a4031b29f97361b541e71438aa7f3ea82212f2

Manually unboost if not using boost in test_codec

2016-05-14T21:11:52+00:00

Right now GUI boost interfers with test_codec running unboosted
by varying the CPU frequency.  This at least makes it more consistent
if run without user input.

Change-Id: I121615098d6bd72ccd13c4e665cc2e4d45d7b58c