mirror of
https://github.com/Telecominfraproject/OpenCellular.git
synced 2025-12-27 18:25:05 +00:00
8a9817a5c7
Refactored the i2c unwedge code to place it in the common directory so that any EC chip can use it. Added to the STM32F and LM4 boards, code to automatically detect and unwedge the i2c bus at the start of an i2c transaction. Note that STM32L already had this ability. To enable unwedging of the i2c port though, the gpio pins for SDA and SCL must be defined in the i2c_ports[] array in the board.c file. This allows the i2c module to bit bang the unwedging for the given port. If SDA and SCL are not defined for the port, then the unwedge code will not run. BUG=chrome-os-partner:26315, chrome-os-partner:23802 BRANCH=none TEST=Manual testing on machines with different EC chips. Testing made extensive use of https://chromium-review.googlesource.com/66389 in order to force wedging of the i2c bus so that we can attempt to unwedge it. Note that you can easily test if the bus is wedged by running i2cscan. On pit and spring: On pit, after each of the following, I verified that the bus was automatically unwedged. On spring, the unwedge only runs at reboot, so, for the non-reboot wedge commands, I manually ran console command unwedge, and verified that the bus became unwedged. (1) Bit bang a transaction but only read part of the response. Command to wedge: i2cwedge 0x90 0 2 2 (2) Bit bang a transaction to do a "write" and stop while the other side is acking. Command to wedge: i2cwedge 0x90 0 1 (3) Same as (1) but do a reboot instead of returning and see that the unwedge works at init time w/ no cancelled transactions. Command to wedge: i2cwedge 0x90 0 6 2 (4) Same as (2) but do a reboot instead of returning and see that the unwedge works at init time w/ no cancelled transactions. Command to wedge: i2cwedge 0x90 0 5 On glimmer: Added code to call i2c_unwedge in accel_init(). Then tested unwedging the accelerometer with the following. One extra difficulty testing this with the accelerometer is that sometimes the bit you stop on is high, which means it won't be wedged at all, the next start transaction will reset the bus. So, sometimes running i2cwedge won't wedge the bus and sometimes it will depending on the acceleration data. (1) Big bang transaction to do a "read" of accelerometer and stop partway: i2cwedge 0x1c 0x0f 2 2 i2cscan to make sure bus is actually wedged i2cunwedge i2cscan to make sure bus is now unwedged. (2) Bit bang transaction to do a "read" and stop partway, then reboot: i2cwedge 0x1c 0x0f 6 2. i2cscan to verify that the bus is working after the reboot. Change-Id: Ie3328e843ffb40f5001c96626fea131c0f9ad9b1 Signed-off-by: Alec Berg <alecaberg@chromium.org> Reviewed-on: https://chromium-review.googlesource.com/188422 Reviewed-by: Randall Spangler <rspangler@chromium.org>
In the most general case, the flash layout looks something like this: +---------------------+ | Reserved for EC use | +---------------------+ +---------------------+ | Vblock B | +---------------------+ | RW firmware B | +---------------------+ +---------------------+ | Vblock A | +---------------------+ | RW firmware A | +---------------------+ +---------------------+ | FMAP | +---------------------+ | Public root key | +---------------------+ | Read-only firmware | +---------------------+ BIOS firmware (and kernel) put the vblock info at the start of each image where it's easy to find. The Blizzard EC expects the firmware vector table to come first, so we have to put the vblock at the end. This means we have to know where to look for it, but that's built into the FMAP and the RO firmware anyway, so that's not an issue. The RO firmware doesn't need a vblock of course, but it does need some reserved space for vboot-related things. Using SHA256/RSA4096, the vblock is 2468 bytes (0x9a4), while the public root key is 1064 bytes (0x428) and the current FMAP is 644 bytes (0x284). If we reserve 4K at the top of each FW image, that should give us plenty of room for vboot-related stuff.