Reserve 1MB of DRAM for loading rockbox and use the rest as BSS
Write sdram setup in assembler and move it to a separate file, together
with MMU init code
git-svn-id: svn://svn.rockbox.org/rockbox/trunk@26926 a1c6a512-1295-4272-9138-f99709370657
not working yet : an empty drive is detected
won't link on fuzev1: drivers allocated in usb_storage.c are too large
git-svn-id: svn://svn.rockbox.org/rockbox/trunk@26905 a1c6a512-1295-4272-9138-f99709370657
has made it more reliable. It seems to basically work now, but
somehow when Linux tries to read the partition table and it gets the
"scsi read10" command, the storage_read_sectors(drive=0, start=0, count=8)
call never returns...
git-svn-id: svn://svn.rockbox.org/rockbox/trunk@26894 a1c6a512-1295-4272-9138-f99709370657
The wakeup alarm will be disabled when powering off anyway
OF database refresh of µSD now only happens if the wakeup was
explicitely enabled
There still seems to be a freeze problem when powering off (sometimes)
git-svn-id: svn://svn.rockbox.org/rockbox/trunk@26875 a1c6a512-1295-4272-9138-f99709370657
Gets a bit more battery life
Adjust current usage, it was a bit too low for Clipv2/Clip+
git-svn-id: svn://svn.rockbox.org/rockbox/trunk@26866 a1c6a512-1295-4272-9138-f99709370657
This fix adds one nop instruction per loop (there were 3 before) and so i2c will be slightly slower, but this hardly a problem.
git-svn-id: svn://svn.rockbox.org/rockbox/trunk@26850 a1c6a512-1295-4272-9138-f99709370657
Once all targets are converted we should probably remove all traces of it entirely.
git-svn-id: svn://svn.rockbox.org/rockbox/trunk@26791 a1c6a512-1295-4272-9138-f99709370657
Fix the asm constraints to mark the left output variable as earlyclobber
Thanks to amiconn
git-svn-id: svn://svn.rockbox.org/rockbox/trunk@26777 a1c6a512-1295-4272-9138-f99709370657
FIQ/IRQ handlers can't be built with -mthumb, and the thumb case was
buggy anyway (spotted by jhMikeS) : pc offset hadn't been adjusted
git-svn-id: svn://svn.rockbox.org/rockbox/trunk@26765 a1c6a512-1295-4272-9138-f99709370657
When building with eabi gcc, 2 variables are assigned the same register
This might be a bug in the constraints, but it's just simpler to use C,
even at the cost of 1 cycle per loop
git-svn-id: svn://svn.rockbox.org/rockbox/trunk@26761 a1c6a512-1295-4272-9138-f99709370657
We can't pop into pc on ARMv4t when using thumb: the T bit won't be
modified if we are returning to a thumb function
Code running on ARMv4t should use the new ldrpc / ldmpc macros instead
of ldr pc, [sp], #4 and ldm(cond) sp!, {regs, pc}
No modification on pure ARM builds and ARMv5+
Note: USE_THUMB is currently never defined, no targets can currently be
built with -mthumb, see FS#6734
git-svn-id: svn://svn.rockbox.org/rockbox/trunk@26756 a1c6a512-1295-4272-9138-f99709370657
preventing the DMA callback to advance in the recording buffer is not
enough in pcm_rec_dma_get_peak_buffer()
we need to prevent it interrupting us while we are copying the left
channel data into the right channel
git-svn-id: svn://svn.rockbox.org/rockbox/trunk@26663 a1c6a512-1295-4272-9138-f99709370657
Fix panics when deleting large files from µSD
Tested on problematic Transcend 4GB Class 6
git-svn-id: svn://svn.rockbox.org/rockbox/trunk@26631 a1c6a512-1295-4272-9138-f99709370657
This adds proper full speed charging support for iPod 4G, Color, Mini 1G, Mini 2G, Nano 1G, and Video. Nano 2G already has support, so now all ipods charge (except 1G-3G whose hardware doesn't support USB charging).
USB charging defaults to on, but will only charge at full speed from a USB host. To charge from an AC charger, you need to set the USB charging mode to "force". It may be possible to detect AC chargers in a future version, at least on some models.
git-svn-id: svn://svn.rockbox.org/rockbox/trunk@26626 a1c6a512-1295-4272-9138-f99709370657
keypad.
Copyright is restored for each keymap according to git blame.
git-svn-id: svn://svn.rockbox.org/rockbox/trunk@26615 a1c6a512-1295-4272-9138-f99709370657
Enable support for the "force" mode of USB charging. This should work on Gigabeat S and Nano2g (and any other future target which has a RB usb stack and supports charging) - if a host connection is not detected within 10 seconds of USB insertion, assume that the connected device is an AC charger and charge anyway, if the user has specified "force" as the mode.
git-svn-id: svn://svn.rockbox.org/rockbox/trunk@26594 a1c6a512-1295-4272-9138-f99709370657
1) "Charge during USB connection" option is now tristate: off/on/force. Currently "force" behaves just like "on", but in future it will allow charging even when it was not possible to positively identify a charger.
2) The H300 code has been adjusted to use the new system but there should be no functional differences, it already had the USB charging option and its USB/charging support is hardware controlled.
3) The Gigabeat S code has been adjusted to use the new system: the player now has the USB charging option, which wasn't previously available. The player will only charge at full speed when allowed to do so by a working USB host, so USB AC adapters won't work very well; however, they didn't work before either, so this is not a change in functionality.
4) The iPod Nano 2G code has been adjusted to use the new system: it already had the USB charging option. Using a USB AC adapter won't charge at full speed any more (it did before) - the old implementation was equivalent to the not-yet-implemented "force" option in the new system.
No other target should be affected. Support for the "force" mode and support for at least some other iPod models will come in a future commit :)
git-svn-id: svn://svn.rockbox.org/rockbox/trunk@26570 a1c6a512-1295-4272-9138-f99709370657
(forward port of r26481 + cosmetics)
This lead to freezes under various conditons (bug report: FS#11267)
Battery life is almost not (if at all) affected:
- between 4 and 8 minutes less runtime for fuzev1
- 9 minutes less runtime for c200v2 (2 minutes difference on a 2 hours run)
Additional verifications of the runtime (including on other models) wouldn't hurt
git-svn-id: svn://svn.rockbox.org/rockbox/trunk@26512 a1c6a512-1295-4272-9138-f99709370657
reorder static functions to avoid the need for early prototypes
panic if the INT_AUDIO callback could not read the 3 IRQ_ENRD registers
git-svn-id: svn://svn.rockbox.org/rockbox/trunk@26445 a1c6a512-1295-4272-9138-f99709370657
this removes workaround for spurious interrupts added in r26316
side effect: make keyclicks work correctly when used intensively (with
the scrollwheel for example), the clicks should be inserted fast enough
for the PCM fifo to not become empty.
PL190 can cause spurious interrupts if an interrupt is triggered by a
peripheral shortly before the interrupt for this peripheral is cleared
from VIC registers.
As an effect we should only write in VIC_INT_EN_CLEAR after having
disabled the source of the interrupt in the peripheral.
references on PL190 spurious interrupts:
- http://www.embeddedrelated.com/groups/lpc2000/show/14599.php
- http://water.cse.unsw.edu.au/esdk/lpc2/spurious-irq.html
git-svn-id: svn://svn.rockbox.org/rockbox/trunk@26444 a1c6a512-1295-4272-9138-f99709370657
This is equivalent to mov pc
Remove an instruction from the asm() for PP cop restart : use register constraint
git-svn-id: svn://svn.rockbox.org/rockbox/trunk@26431 a1c6a512-1295-4272-9138-f99709370657
Skipped the specifiers for octal, (u)int_leastN_t, (u)int_fastN_t and
(u)intmax_t as we don't use them in rockbox
Specifiers use the types specified in stdint.h
Specifiers for short (16 bits) and char (8 bits) use the int specifier,
our format.c doesn't know about 'h' and 'hh'
git-svn-id: svn://svn.rockbox.org/rockbox/trunk@26411 a1c6a512-1295-4272-9138-f99709370657
Using BX has the effect to set the T bit, so it can return to a thumb
function
With this change, rockbox runs fine on Clipv2 built with -mthumb (for
most files which don't use inline 32 bits ARM assembly)
Some other places use code which change the T bit on armv5 but not on
armv4 so armv4 won't run
See FS#6734
git-svn-id: svn://svn.rockbox.org/rockbox/trunk@26386 a1c6a512-1295-4272-9138-f99709370657
invert setting isn't remembered when backlight goes off and lcd_enable()
is ran again
git-svn-id: svn://svn.rockbox.org/rockbox/trunk@26374 a1c6a512-1295-4272-9138-f99709370657
Waiting for TRAN state after the end of the transfer caused writing problems for some uSD cards. Removing this wait returns write functionality to those cards.
git-svn-id: svn://svn.rockbox.org/rockbox/trunk@26365 a1c6a512-1295-4272-9138-f99709370657
(apply r23738 to as3525v2 code)
extend the arbitrary delay to 5 seconds, we never know
git-svn-id: svn://svn.rockbox.org/rockbox/trunk@26342 a1c6a512-1295-4272-9138-f99709370657
tick_start, fmradio_i2c_init, pcm_init, pcm_postinit, pcm_play_dma_init
Also add some as3525 target specific functions
Make as3525v2 use this attribute
git-svn-id: svn://svn.rockbox.org/rockbox/trunk@26328 a1c6a512-1295-4272-9138-f99709370657
It's now easier to force rebuild of files depending on the svn revision
version.c/version.h are generated once with new tools/genversion.sh
Changes in the VCS are still not auto detected, so you'll have to remove
builddir/version.* if you want to change the string in your binaries
APPSVERSION is now called RBVERSION and is defined in the generated
header instead of being defined by the Makefiles
appsversion is now called rbversion (the plugin api number didn't change
since old modules are still binary compatible)
Change some bootloaders to use knwon-at-buildtime RBVERSION instead of
"%s" + rbversion
You'll need to run make clean to regenerate dependencies after the
removal of apps/version.h
To build binaries with a different version string, hand-edit
tools/version.sh or tools/genversion.sh (which calls the former)
git-svn-id: svn://svn.rockbox.org/rockbox/trunk@26320 a1c6a512-1295-4272-9138-f99709370657
Even if we disable it in VIC_INT_EN_CLEAR, it still fires an interrupt
by running the default isr.
Locking requires disabling interrupts completely.
I'm not sure if DMAC is the only interrupt affected or if we can't rely
on VIC_INT_EN_CLEAR to mask interrupts.
If it's the latter, we will need to have special locking scheme for
INT_AUDIO.
git-svn-id: svn://svn.rockbox.org/rockbox/trunk@26316 a1c6a512-1295-4272-9138-f99709370657
since it's now the default thread which is implicitely created by starting the sim.
This caused a segfault that r26283 tried to fix. Revert r26283 and protect
the main thread from being removed.
git-svn-id: svn://svn.rockbox.org/rockbox/trunk@26315 a1c6a512-1295-4272-9138-f99709370657
Declare VIC registers holding function pointers as volatile pointers to
function pointers and access them directly without casting
UIRQ() is an IRQ handler too, even if it doesn't return
git-svn-id: svn://svn.rockbox.org/rockbox/trunk@26313 a1c6a512-1295-4272-9138-f99709370657
- retry failed commands up to 6 times
- wait at most 5 seconds for the card to go to TRAN state
- wait for the card to be in TRAN state (i.e. not programming) before
disabling SD clocks (remove the workaround in sd_enable() )
should fix lockups reported on fuzev1 under heavy internal storage + µSD use
(playback + database/pictureflow updating)
git-svn-id: svn://svn.rockbox.org/rockbox/trunk@26291 a1c6a512-1295-4272-9138-f99709370657
Selecting the right card before the transfer will deselect the other one
git-svn-id: svn://svn.rockbox.org/rockbox/trunk@26290 a1c6a512-1295-4272-9138-f99709370657
We assume only one bit is set and use clz to find the most significant
set bit, if there's more than one bit set we'll see it in the panic msg
git-svn-id: svn://svn.rockbox.org/rockbox/trunk@26285 a1c6a512-1295-4272-9138-f99709370657
If we disable the hearbeat source after enabling the watchdog there's no
more lock-up when powering off the dap
Comment out alarm enabling on startup since rockbox disables the alarm anyway
git-svn-id: svn://svn.rockbox.org/rockbox/trunk@26247 a1c6a512-1295-4272-9138-f99709370657
A specific poweroff function needs to be used for wake-up to work
Disable RTC in bootloaders for consistency with other Sansas
git-svn-id: svn://svn.rockbox.org/rockbox/trunk@26243 a1c6a512-1295-4272-9138-f99709370657
- fix check_event_queue to use queue_peek and not mess message order
- change check_dircache_state to a more modest check to avoid messing up dircache state
git-svn-id: svn://svn.rockbox.org/rockbox/trunk@26228 a1c6a512-1295-4272-9138-f99709370657
Charge using 150mA current like the OF to charge faster
Edit clipv2 charging curve with 150mA
git-svn-id: svn://svn.rockbox.org/rockbox/trunk@26227 a1c6a512-1295-4272-9138-f99709370657
Move code not defined for bootloader under a single preprocessor check
git-svn-id: svn://svn.rockbox.org/rockbox/trunk@26219 a1c6a512-1295-4272-9138-f99709370657
GPIO direction needs only to be set at init
both functions are now one liners so make them static inline
git-svn-id: svn://svn.rockbox.org/rockbox/trunk@26218 a1c6a512-1295-4272-9138-f99709370657
It seems there is some significant variance between different greyscale ipods of the same model, so it's unlikely that *any* value will be great for everyone. I'm hoping that this is just high enough that the lightest screens aren't totally blank, but not dark enough that the darkest screens are entirely unreadably black ;)
This rather unsatisfactory compromise is to help us get a new bootloader release out (since bootloaders can't have their contrast adjusted) - a real fix would be to detect the LCD type, which the OF appears to be able to do (maybe).
git-svn-id: svn://svn.rockbox.org/rockbox/trunk@26208 a1c6a512-1295-4272-9138-f99709370657
- forgot a ccu undocumented thing in init
- finish init code (following linux patch)
- enable interrupts
- don't use dynanmic fifo sizing by default
- move to a more static configuration of usb hardware parameters (it might be necessary to move them to config* files later).
git-svn-id: svn://svn.rockbox.org/rockbox/trunk@26195 a1c6a512-1295-4272-9138-f99709370657
it might be valid C but is confusing, and not consistent with the rest of file.c
git-svn-id: svn://svn.rockbox.org/rockbox/trunk@26191 a1c6a512-1295-4272-9138-f99709370657
we still need to convert uncached addresses to physical addresses
git-svn-id: svn://svn.rockbox.org/rockbox/trunk@26177 a1c6a512-1295-4272-9138-f99709370657
No need for special address handling, all addresses are equal to their physical address
git-svn-id: svn://svn.rockbox.org/rockbox/trunk@26176 a1c6a512-1295-4272-9138-f99709370657
- enhance registers description in header file with respect to the as3543
registers: point which one are present or not in the as3543, which bits
change (i didn't detail all registers)
- cache less registers, only up to the ones which are written to
multiple times
- use ascodec_readbytes() to cache all needed registers in a row
- separate the as3543 bits from as3514/as3515 in audiohw_preinit()
- comment out unused audiohw_set_lineout_vol
- some cosmetics
git-svn-id: svn://svn.rockbox.org/rockbox/trunk@26174 a1c6a512-1295-4272-9138-f99709370657
it's only needed if we hotswap the µSD or if we build the fuzev2
scrollwheel code without irq (which we don't do anyway)
git-svn-id: svn://svn.rockbox.org/rockbox/trunk@26165 a1c6a512-1295-4272-9138-f99709370657
test_disk doesn't freeze anymore when closing the file between writing & reading
git-svn-id: svn://svn.rockbox.org/rockbox/trunk@26164 a1c6a512-1295-4272-9138-f99709370657
The list is ordered by priority, put the DMA and SD interrupts at
the top
This seems to fix the random freezes on heavy storage operation like
building database or pictureflow cache
The recently added audio interrupt had a priority higher than SD slot on
as3525v1, give it a lower priority
git-svn-id: svn://svn.rockbox.org/rockbox/trunk@26160 a1c6a512-1295-4272-9138-f99709370657
