With the bringup patches, *usb start* leads to data abort in free() back trace below. The backtrace doesn't involve any of the code added in this series, but still I see a crash. Has anyone faced similar kind of issue? I see samsung xhci drivers is already converted to DM, was there any similar issue faced?
When I hacked free() to return without doing anything it passes and I am able to load kernel from a usb drive. Pushed a branch [1] if any one interested in testing, my test logs [2]
free(void *)() at dlmalloc.c:2,462 OxBFF702F2 free_pipelist(struct pipe*, int)() at cli_hush.c:1,986 OxBFF62794 parse_stream_outer(struct in_str *, int)() at cli_hush.c:3,197 OxBFF63904 parsefile_outer() at cli_hush.c:3,281 OxBFF6397A cli loop() at cli.c:207 OxBFF71A96 main_loop() at main.c:89 OxBFF62442 run_main_loop() at board_r.c:731 OxBFF63DB2 initcall_run_list(int (*)() *)() at initcall.c:32 OxBFF87E5E board_init_r(struct global_data *, unsigned long)() at board_r.c:976 OxBFF63EB8 OxBFF6194C (no symbols are defined for OxBFF6194C)
[1]: git://git.ti.com/~mugunthanvnm/ti-u-boot/mugunth-ti-u-boot.git dm-usb-rfc [2]: http://pastebin.ubuntu.com/14398751/
Mugunthan V N (4): configs: am43xx: enabled USB support for U-boot drivers: usb: xhci-omap: convert driver to adopt driver model arm: dts: am4372: add xhci-omap compatible to host node defconfig: am437x_sk_evm: enable usb driver model
arch/arm/dts/am4372.dtsi | 2 +- configs/am437x_sk_evm_defconfig | 2 ++ drivers/usb/host/xhci-omap.c | 77 +++++++++++++++++++++++++++++++++++++++++ include/configs/am43xx_evm.h | 3 +- 4 files changed, 82 insertions(+), 2 deletions(-)