Re: [PATCH v8 03/14] x86: Add support for function granular KASLR

[Peter Zijlstra]

On Thu, Dec 02, 2021 at 11:32:03PM +0100, Alexander Lobakin wrote:
> +static struct orc_entry *cur_orc_table;
> +static int *cur_orc_ip_table;

> +static int cmp_section_addr_orc(const void *a, const void *b)
> +{
> +	unsigned long ptr = (unsigned long)a;
> +	Elf_Shdr *s = *(Elf_Shdr **)b;
> +	unsigned long end = s->sh_addr + s->sh_size;
> +
> +	/* orc relocations can be one past the end of the section */
> +	if (ptr >= s->sh_addr && ptr <= end)
> +		return 0;
> +
> +	if (ptr < s->sh_addr)
> +		return -1;
> +
> +	return 1;
> +}
> +
> +/*
> + * Discover if the orc_unwind address is in a randomized section and if so,
> + * adjust by the saved offset.
> + */
> +Elf_Shdr *adjust_address_orc(long *address)
> +{
> +	Elf_Shdr **s;
> +	Elf_Shdr *shdr;
> +
> +	if (nofgkaslr)
> +		return NULL;
> +
> +	s = bsearch((const void *)*address, sections, sections_size, sizeof(*s),
> +		    cmp_section_addr_orc);
> +	if (s) {
> +		shdr = *s;
> +		*address += shdr->sh_offset;
> +		return shdr;
> +	}
> +
> +	return NULL;
> +}

> +static inline unsigned long orc_ip(const int *ip)
> +{
> +	return (unsigned long)ip + *ip;
> +}
> +
> +static void orc_sort_swap(void *_a, void *_b, int size)
> +{
> +	struct orc_entry *orc_a, *orc_b;
> +	struct orc_entry orc_tmp;
> +	int *a = _a, *b = _b, tmp;
> +	int delta = _b - _a;
> +
> +	/* Swap the .orc_unwind_ip entries: */
> +	tmp = *a;
> +	*a = *b + delta;
> +	*b = tmp - delta;
> +
> +	/* Swap the corresponding .orc_unwind entries: */
> +	orc_a = cur_orc_table + (a - cur_orc_ip_table);
> +	orc_b = cur_orc_table + (b - cur_orc_ip_table);
> +	orc_tmp = *orc_a;
> +	*orc_a = *orc_b;
> +	*orc_b = orc_tmp;
> +}
> +
> +static int orc_sort_cmp(const void *_a, const void *_b)
> +{
> +	struct orc_entry *orc_a;
> +	const int *a = _a, *b = _b;
> +	unsigned long a_val = orc_ip(a);
> +	unsigned long b_val = orc_ip(b);
> +
> +	if (a_val > b_val)
> +		return 1;
> +	if (a_val < b_val)
> +		return -1;
> +
> +	/*
> +	 * The "weak" section terminator entries need to always be on the left
> +	 * to ensure the lookup code skips them in favor of real entries.
> +	 * These terminator entries exist to handle any gaps created by
> +	 * whitelisted .o files which didn't get objtool generation.
> +	 */
> +	orc_a = cur_orc_table + (a - cur_orc_ip_table);
> +	return orc_a->sp_reg == ORC_REG_UNDEFINED && !orc_a->end ? -1 : 1;
> +}
> +
> +static void update_orc_table(unsigned long map)
> +{
> +	int i;
> +	int num_entries =
> +		(addr___stop_orc_unwind_ip - addr___start_orc_unwind_ip) / sizeof(int);
> +
> +	cur_orc_ip_table = (int *)(addr___start_orc_unwind_ip + map);
> +	cur_orc_table = (struct orc_entry *)(addr___start_orc_unwind + map);
> +
> +	debug_putstr("\nUpdating orc tables...\n");
> +	for (i = 0; i < num_entries; i++) {
> +		unsigned long ip = orc_ip(&cur_orc_ip_table[i]);
> +		Elf_Shdr *s;
> +
> +		/* check each address to see if it needs adjusting */
> +		ip = ip - map;
> +
> +		/*
> +		 * objtool places terminator entries just outside the end of
> +		 * the section. To identify an orc_unwind_ip address that might
> +		 * need adjusting, the address should be compared differently
> +		 * than a normal address.
> +		 */
> +		s = adjust_address_orc(&ip);
> +		if (s)
> +			cur_orc_ip_table[i] += s->sh_offset;
> +	}
> +}
> +
> +static void sort_orc_table(unsigned long map)
> +{
> +	int num_entries =
> +		(addr___stop_orc_unwind_ip - addr___start_orc_unwind_ip) / sizeof(int);
> +
> +	cur_orc_ip_table = (int *)(addr___start_orc_unwind_ip + map);
> +	cur_orc_table = (struct orc_entry *)(addr___start_orc_unwind + map);
> +
> +	debug_putstr("\nRe-sorting orc tables...\n");
> +	sort(cur_orc_ip_table, num_entries, sizeof(int), orc_sort_cmp,
> +	     orc_sort_swap);
> +}

Is this somehow different from what we already have in
arch/x86/kernel/unwind_orc.c for module support? Do we really need two
copies of all that?