bpf: hardcode BPF_PROG_PACK_SIZE to 2MB * num_possible_nodes()

On some architectures like ARM64, PMD_SIZE can be really large in some
configurations. Like with CONFIG_ARM64_64K_PAGES=y the PMD_SIZE is
512MB.

Use 2MB * num_possible_nodes() as the size for allocations done through
the prog pack allocator. On most architectures, PMD_SIZE will be equal
to 2MB in case of 4KB pages and will be greater than 2MB for bigger page
sizes.

Fixes: ea2baba ("bpf: Simplify bpf_prog_pack_[size|mask]")
Reported-by: "kernelci.org bot" <[email protected]>
Closes: https://lore.kernel.org/all/[email protected]/
Reported-by: kernel test robot <[email protected]>
Closes: https://lore.kernel.org/oe-kbuild-all/[email protected]/
Suggested-by: Song Liu <[email protected]>
Signed-off-by: Puranjay Mohan <[email protected]>
Message-ID: <[email protected]>
Signed-off-by: Alexei Starovoitov <[email protected]>

Author: puranjaymohan

kernel/bpf/core.c

@@ -893,7 +893,12 @@ static LIST_HEAD(pack_list);
  * CONFIG_MMU=n. Use PAGE_SIZE in these cases.
  */
 #ifdef PMD_SIZE
-#define BPF_PROG_PACK_SIZE (PMD_SIZE * num_possible_nodes())
+/* PMD_SIZE is really big for some archs. It doesn't make sense to
+ * reserve too much memory in one allocation. Hardcode BPF_PROG_PACK_SIZE to
+ * 2MiB * num_possible_nodes(). On most architectures PMD_SIZE will be
+ * greater than or equal to 2MB.
+ */
+#define BPF_PROG_PACK_SIZE (SZ_2M * num_possible_nodes())
 #else
 #define BPF_PROG_PACK_SIZE PAGE_SIZE
 #endif