Re: [PATCH v2 1/2] tools/vm/page_owner_sort.c: Sort by stacktrace before culling

[Sean Anderson]

On 11/25/21 4:46 AM, weizhenliang wrote:
> On 2021/11/25 3:37, Sean Anderson <seanga2@xxxxxxxxx> wrote:
>
> > static int compare_num(const void *p1, const void *p2) @@ -121,6
> > +122,7 @@ static void add_list(char *buf, int len)
> > list[list_size].page_num = get_page_num(buf);
> > memcpy(list[list_size].txt, buf, len);  list[list_size].txt[len] = 0;
> > + list[list_size].stacktrace = strchr(list[list_size].txt, '\n');
>
> When read_block gets an empty line, buf is "\n", then the stacktrace is NULL
>
> > list_size++;
> > if (list_size % 1000 == 0) {
> > printf("loaded %d\r", list_size);
> > @@ -199,7 +201,7 @@ int main(int argc, char **argv)
> >
> > printf("sorting ....\n");
> >
> > - qsort(list, list_size, sizeof(list[0]), compare_txt);
> > + qsort(list, list_size, sizeof(list[0]), compare_stacktrace);
> >
> > list2 = malloc(sizeof(*list) * list_size);  if (!list2) { @@ -211,7
> > +213,7 @@ int main(int argc, char **argv)
> >
> > for (i = count = 0; i < list_size; i++) {  if (count == 0 ||
> > - strcmp(list2[count-1].txt, list[i].txt) != 0) {
> > + strcmp(list2[count-1].stacktrace, list[i].stacktrace) != 0) {
>
> And when stacktrace is NULL, a segmentation fault will be triggered here.

Ah, whoops. Looks like I check for this in compare_stacktrace but not here.

> > list2[count++] = list[i];
> > } else {
> > list2[count-1].num += list[i].num;
>
> 1. Maybe you can check whether the ret of read_block is 0 before add_list,
> or whether the len of buf is 0 in add_list

I think this is the best route.

Since this seems to have already been applied I've sent a follow-up patch.

--Sean