On Tue, Jan 21, 2014 at 02:49:35AM +0000, Anton Blanchard wrote:
>
> The sort option in dtc treats unit addresses as strings. This causes
> cpu nodes to end up out of order:
>
> # dtc -s -I fs -O dts /proc/device-tree | grep PowerPC,POWER7
>
> PowerPC,POWER7@30 {
> PowerPC,POWER7@68 {
> PowerPC,POWER7@70 {
> PowerPC,POWER7@828 {
> PowerPC,POWER7@860 {
> PowerPC,POWER7@868 {
> PowerPC,POWER7@8a0 {
> PowerPC,POWER7@8b0 {
> PowerPC,POWER7@8f0 {
> PowerPC,POWER7@a0 {
> PowerPC,POWER7@a8 {
> PowerPC,POWER7@e0 {
>
> If we use this device tree for a kexec boot we end up with a confusing
> layout of logical CPUs:
>
> node 0 cpus: 0-23 72-95
> node 0 size: 32633 MB
>
> node 1 cpus: 24-71
> node 1 size: 32631 MB
>
> The reason for this is that we allocate logical CPU ids as we walk
> through the device tree.
>
> In cmp_subnode, if both nodes have a hex unit address and the
> basenames match, then compare by number.
>
> This fixes the issue:
>
> # dtc -s -I fs -O dts /proc/device-tree | grep PowerPC,POWER7
> PowerPC,POWER7@30 {
> PowerPC,POWER7@68 {
> PowerPC,POWER7@70 {
> PowerPC,POWER7@a0 {
> PowerPC,POWER7@a8 {
> PowerPC,POWER7@e0 {
> PowerPC,POWER7@828 {
> PowerPC,POWER7@860 {
> PowerPC,POWER7@868 {
> PowerPC,POWER7@8a0 {
> PowerPC,POWER7@8b0 {
> PowerPC,POWER7@8f0 {
>
> And the CPU layout is as expected:
>
> node 0 cpus: 0-47
> node 0 size: 32633 MB
>
> node 1 cpus: 48-95
> node 1 size: 32631 MB
>
> Signed-off-by: Anton Blanchard <[email protected]>
> --
>
> Index: b/livetree.c
> ===================================================================
> --- a/livetree.c
> +++ b/livetree.c
> @@ -656,12 +656,38 @@ static void sort_properties(struct node
> free(tbl);
> }
>
> +static bool is_hex(const char *str)
> +{
> + while (*str) {
> + if (!isxdigit(*str++))
> + return false;
> + }
> +
> + return true;
> +}
> +
> static int cmp_subnode(const void *ax, const void *bx)
> {
> - const struct node *a, *b;
> + struct node *a, *b;
> + const char *a_unit, *b_unit;
> +
> + a = *((struct node * const *)ax);
> + b = *((struct node * const *)bx);
> +
> + a_unit = get_unitname(a);
> + b_unit = get_unitname(b);
> +
> + /* Sort hex unit addresses by number */
> + if (a_unit && b_unit && (a->basenamelen == b->basenamelen) &&
> + !strncmp(a->name, b->name, a->basenamelen) &&
> + is_hex(a_unit) && is_hex(b_unit)) {
> + unsigned long long a_num, b_num;
> +
> + a_num = strtoull(a_unit, NULL, 16);
> + b_num = strtoull(b_unit, NULL, 16);
>
> - a = *((const struct node * const *)ax);
> - b = *((const struct node * const *)bx);
> + return (a_num > b_num) - (a_num < b_num);
> + }
>
> return strcmp(a->name, b->name);
> }
Minor issue, but when #address-cells == 2, some unit addresses are split
in the middle by a ',' to separate the value of each cell, e.g.
"flash@2,0". For those, is_hex will return false and we'll compare
unit-addresses as strings.
I took a quick look over the dts in the Linux kernel tree (with `git
grep "@.\+," -- arch/*/boot/dts` and I think every instance there would
sort correctly as a string, but it would be nice to fix the issue
regardless of how large the unit-address is.
Perhaps we could have a helper function for reading the unit-address
that would take this into account?
Cheers,
Mark.
--
To unsubscribe from this list: send the line "unsubscribe devicetree" in
the body of a message to [email protected]
More majordomo info at http://vger.kernel.org/majordomo-info.html
