The RPU pytest introduced by commit e894c10c040b ("test/py: zynqmp_rpu: Add test for loading RPU apps") expects 3rd parameter as string not a number that's why extend command to actually handle both. The issue with existing code is that when any non number string is passed hextoul returns 0. For backward compatibility zynqmp tcminit 0/1 can be still used but it is recommended to use strings instead.
Signed-off-by: Padmarao Begari padmarao.begari@amd.com --- arch/arm/mach-zynqmp/zynqmp.c | 16 +++++++--------- 1 file changed, 7 insertions(+), 9 deletions(-)
diff --git a/arch/arm/mach-zynqmp/zynqmp.c b/arch/arm/mach-zynqmp/zynqmp.c index bf39c5472e..bea29bb4c6 100644 --- a/arch/arm/mach-zynqmp/zynqmp.c +++ b/arch/arm/mach-zynqmp/zynqmp.c @@ -151,14 +151,12 @@ static int do_zynqmp_tcm_init(struct cmd_tbl *cmdtp, int flag, int argc, if (argc != cmdtp->maxargs) return CMD_RET_USAGE;
- if (strcmp(argv[2], "lockstep") && strcmp(argv[2], "split")) { - printf("mode param should be lockstep or split\n"); - return CMD_RET_FAILURE; - } - - mode = hextoul(argv[2], NULL); - if (mode != TCM_LOCK && mode != TCM_SPLIT) { - printf("Mode should be either 0(lock)/1(split)\n"); + if (!strcmp(argv[2], "lockstep") || !strcmp(argv[2], "0")) { + mode = TCM_LOCK; + } else if (!strcmp(argv[2], "split") || !strcmp(argv[2], "1")) { + mode = TCM_SPLIT; + } else { + printf("Mode should be either lockstep/split\n"); return CMD_RET_FAILURE; }
@@ -403,7 +401,7 @@ U_BOOT_LONGHELP(zynqmp, " initialized before accessing to avoid ECC\n" " errors. mode specifies in which mode TCM has\n" " to be initialized. Supported modes will be\n" - " lock(0)/split(1)\n" + " lockstep(0)/split(1)\n" #endif "zynqmp pmufw address size - load PMU FW configuration object\n" "zynqmp pmufw node <id> - load PMU FW configuration object, <id> in dec\n"