Reimplement voltage scaling on AMS Sansas at 1.10v during unboosted operation to improve runtimes. The voltage is now also boosted during disk access if a µSD is present. This prevents the µSD problems we saw on the last implementation.
git-svn-id: svn://svn.rockbox.org/rockbox/trunk@23193 a1c6a512-1295-4272-9138-f99709370657
Don't implement usb_enable() is HAVE_USBSTACK is not defined
git-svn-id: svn://svn.rockbox.org/rockbox/trunk@23185 a1c6a512-1295-4272-9138-f99709370657
Implement usb_enable()
Reorder/Modify usb_drv_init() to match closer the OF and remove a freeze
(USB registers were accessed before the USB module was enabled)
Add a panic in USB isr to be sure we notice when it's called
Reset GPIOA direction for usb_detect() to notice extraction
Add some comments
git-svn-id: svn://svn.rockbox.org/rockbox/trunk@23184 a1c6a512-1295-4272-9138-f99709370657
Remove useless E200V2 || FUZE || C200V2 : all those models have
MULTIDRIVE and/or HOTSWAP defined and have no specific difference.
Correct some comments
Remove HAVE_MULTIDRIVE within HAVE_HOTSWAP since HOTSWAP imply MULTIDRIVE
Change HOTSWAP to MULTIDRIVE where needed
Use NUM_DRIVES in sd_num_drives()
git-svn-id: svn://svn.rockbox.org/rockbox/trunk@23180 a1c6a512-1295-4272-9138-f99709370657
The bits which are not divider (i2si/i2so enable and clock selection)
would be unset.
This is not a problem in the current code since this function was always
called before starting playback (and setting those bits) but this might
be a problem when recording is enabled.
Finally it is simpler to read.
Thanks to Fred Bauer for pointing this.
git-svn-id: svn://svn.rockbox.org/rockbox/trunk@23156 a1c6a512-1295-4272-9138-f99709370657
We are unable to successfully put the pl180 controller into 4 bit mode so we should not put the cards into widebus mode for now.
The blocklength is hardcoded to 512 in sd.c and BLOCKLEN defaults to 512 so this command is not needed.
It appears the internal SD card is not HS capable but sending it the HS switch command does not seem to hinder it's init process.
Assume all SD_V2 cards are HS capable and send them the HS switch command.
If View disk info shows 50.0 MBit/s the card has HS timings.
git-svn-id: svn://svn.rockbox.org/rockbox/trunk@23137 a1c6a512-1295-4272-9138-f99709370657
This still doesn't solve the problem of the cards being overclocked at 62MHz but we can mitigate this a bit by switching to HS timings.
git-svn-id: svn://svn.rockbox.org/rockbox/trunk@23088 a1c6a512-1295-4272-9138-f99709370657
of 400 khz Ident freq. Also misc formatting changes.
Displaying the 400 khz ident frequency for the SD and uSD cards was not really
useful information. This change displays the MCICLK frequency that we are
running the cards at. The page now displays 0MHz until a card access and then
will displays the set frequency and the actual frequency. The uSD display is
now only displayed for those players with a uSD
git-svn-id: svn://svn.rockbox.org/rockbox/trunk@22684 a1c6a512-1295-4272-9138-f99709370657
The blue bars I previously had with this way are fixed by blocking DBOP output while reading buttons from GPIO.
We reached the maximum fps even at unboosted now (100.0 fps at lcd_update(), DBOP can't do more). Optimisation to YUV stuff is still possible though.
The e200v2 scrollwheel is very improved due to the this and the previous speedup since the chance of getting the wheel read blocked is greatly decreased with faster updates.
git-svn-id: svn://svn.rockbox.org/rockbox/trunk@22578 a1c6a512-1295-4272-9138-f99709370657
Unaligned (to 4byte boundary) and odd (where the amount of data is not an integer multiple of 32bit) are still supported. Even those still have a speedup of >2.
git-svn-id: svn://svn.rockbox.org/rockbox/trunk@22574 a1c6a512-1295-4272-9138-f99709370657
They would previously appear like coming from the (disabled) watchdog
module
git-svn-id: svn://svn.rockbox.org/rockbox/trunk@22251 a1c6a512-1295-4272-9138-f99709370657
Nothing should change, since the value of transfer didn't change in the == case, but it saves executing this case at least.
git-svn-id: svn://svn.rockbox.org/rockbox/trunk@21931 a1c6a512-1295-4272-9138-f99709370657
The maximum number of errors is 10 (arbitrary)
A recovery mechanism is not in place (yet) but could be implemented in the future
git-svn-id: svn://svn.rockbox.org/rockbox/trunk@21776 a1c6a512-1295-4272-9138-f99709370657
clocks disabling and mutex unlocking were not made in case of errors
git-svn-id: svn://svn.rockbox.org/rockbox/trunk@21773 a1c6a512-1295-4272-9138-f99709370657
If µSD init fails, rockbox will panic and give an error number
Use a maximal delay of 1 second for µSD (and internal storage) init, as
specified in the SD Specification 2.00
git-svn-id: svn://svn.rockbox.org/rockbox/trunk@21742 a1c6a512-1295-4272-9138-f99709370657
Work based on AMD 5536 linux driver and AMD Geode CS5536 datasheet
Rockbox still doesn't reboot on USB insert
git-svn-id: svn://svn.rockbox.org/rockbox/trunk@21689 a1c6a512-1295-4272-9138-f99709370657
