* Huaixin Chang <changhuai...@linux.alibaba.com> wrote:

> Since orc tables are already sorted by sorttable tool, let us move
> building of fast lookup table into sorttable tool too. This saves us
> 6380us from boot time under Intel(R) Xeon(R) CPU E5-2682 v4 @ 2.50GHz
> with 64 cores.

Neat!

> +struct orc_sort_param {
> +     size_t          lookup_table_size;
> +     unsigned int    *orc_lookup_table;
> +     unsigned long   start_ip;
> +     size_t          text_size;
> +     unsigned int    orc_num_entries;
> +};

>  
> +#define LOOKUP_BLOCK_ORDER   8
> +#define LOOKUP_BLOCK_SIZE    (1 << LOOKUP_BLOCK_ORDER)
> +
> +     for (i = 0; i < lookup_num_blocks-1; i++) {
> +             orc = __orc_find(g_orc_ip_table, g_orc_table,
> +                              num_entries,
> +                              lookup_start_ip + (LOOKUP_BLOCK_SIZE * i));
> +             if (!orc) {
> +                     snprintf(g_err, ERRSTR_MAXSZ,
> +                                     "Corrupt .orc_unwind table\n");
> +                     pthread_exit(g_err);
> +             }
> +
> +             orc_lookup[i] = orc - g_orc_table;
> +     }
> +
> +     /* Initialize the ending block: */
> +     orc = __orc_find(g_orc_ip_table, g_orc_table, num_entries,
> +                      lookup_stop_ip);
> +     if (!orc) {
> +             snprintf(g_err, ERRSTR_MAXSZ, "Corrupt .orc_unwind table\n");
> +             pthread_exit(g_err);
> +     }
> +     orc_lookup[lookup_num_blocks-1] = orc - g_orc_table;

Yeah, so now this definition of LOOKUP_BLOCK_* basicaly duplicates the 
arch/x86/include/asm/orc_lookup.h size, with no obvious link between 
the two. This is asking for trouble.

<asm/orc_lookup.h> looks simple enough - can we include it in 
scripts/sorttable.h?

Or better yet, please move these two defines into <asm/orc_types.h>, 
which is already included in sorttable.h.

BTW., please update your patches to spell 'ORC' in a capitalized 
fashion, like most of the existing code does:

>       /* create thread to sort ORC unwind tables concurrently */

Thanks,

        Ingo

Reply via email to