raindrops RubyGem user+dev discussion/patches/pulls/bugs/help
 help / color / mirror / code / Atom feed
From: Eric Wong <bofh@yhbt.net>
To: raindrops-public@yhbt.net
Subject: [PATCH 3/4] linux_inet_diag: get rid of alloca usage
Date: Sun, 11 Jun 2023 21:33:27 +0000	[thread overview]
Message-ID: <20230611213328.379546-4-bofh@yhbt.net> (raw)
In-Reply-To: <20230611213328.379546-1-bofh@yhbt.net>

alloca makes stack usage unpredictable and life difficult for
static analysis tools and compilers.  The 46 bytes of
INET6_ADDRSTRLEN is fine to keep on stack, but page size can be
several MB large in some architectures (but typically 4K on
common architectures).

Thus we handle page size-ed allocations via `rb_str_tmp_new'.
`rb_str_tmp_new' has been in public Ruby headers since the 1.9
days and used by the core `zlib', `digest', and `zlib'
extensions, so it should be safe to use (and `rb_str_resize' is
used in many more C extensions).
---
 ext/raindrops/linux_inet_diag.c | 19 +++++++++++--------
 1 file changed, 11 insertions(+), 8 deletions(-)

diff --git a/ext/raindrops/linux_inet_diag.c b/ext/raindrops/linux_inet_diag.c
index e1ae62a..2d4f503 100644
--- a/ext/raindrops/linux_inet_diag.c
+++ b/ext/raindrops/linux_inet_diag.c
@@ -212,24 +212,25 @@ static void bug_warn_nogvl(const char *fmt, ...)
 static struct listen_stats *stats_for(st_table *table, struct inet_diag_msg *r)
 {
 	char *host, *key, *port, *old_key;
-	size_t alloca_len;
 	struct listen_stats *stats;
 	socklen_t hostlen;
 	socklen_t portlen = (socklen_t)sizeof("65535");
 	int n;
 	const void *src = r->id.idiag_src;
+	char buf[INET6_ADDRSTRLEN];
+	size_t buf_len;
 
 	switch (r->idiag_family) {
 	case AF_INET: {
 		hostlen = INET_ADDRSTRLEN;
-		alloca_len = hostlen + portlen;
-		host = key = alloca(alloca_len);
+		buf_len = hostlen + portlen;
+		host = key = buf;
 		break;
 		}
 	case AF_INET6: {
 		hostlen = INET6_ADDRSTRLEN;
-		alloca_len = 1 + hostlen + 1 + portlen;
-		key = alloca(alloca_len);
+		buf_len = 1 + hostlen + 1 + portlen;
+		key = buf;
 		host = key + 1;
 		break;
 		}
@@ -269,7 +270,7 @@ static struct listen_stats *stats_for(st_table *table, struct inet_diag_msg *r)
 	old_key = key;
 
 	if (r->idiag_state == TCP_ESTABLISHED) {
-		n = snprintf(key, alloca_len, "%s:%u",
+		n = snprintf(key, buf_len, "%s:%u",
 				 addr_any(r->idiag_family),
 				 ntohs(r->id.idiag_sport));
 		if (n <= 0) {
@@ -615,7 +616,7 @@ static VALUE tcp_listener_stats(int argc, VALUE *argv, VALUE self)
 {
 	VALUE rv = rb_hash_new();
 	struct nogvl_args args;
-	VALUE addrs, sock;
+	VALUE addrs, sock, buf;
 
 	rb_scan_args(argc, argv, "02", &addrs, &sock);
 
@@ -624,8 +625,9 @@ static VALUE tcp_listener_stats(int argc, VALUE *argv, VALUE self)
 	 * buffer for recvmsg() later, we already checked for
 	 * OPLEN <= page_size at initialization
 	 */
+	buf = rb_str_buf_new(page_size);
 	args.iov[2].iov_len = OPLEN;
-	args.iov[2].iov_base = alloca(page_size);
+	args.iov[2].iov_base = RSTRING_PTR(buf);
 	args.table = NULL;
 	sock = NIL_P(sock) ? rb_funcall(cIDSock, id_new, 0)
 			: rb_io_get_io(sock);
@@ -672,6 +674,7 @@ static VALUE tcp_listener_stats(int argc, VALUE *argv, VALUE self)
 		rb_hash_foreach(rv, drop_placeholders, Qfalse);
 
 	/* let GC deal with corner cases */
+	rb_str_resize(buf, 0);
 	if (argc < 2) rb_io_close(sock);
 	return rv;
 }

  parent reply	other threads:[~2023-06-11 21:33 UTC|newest]

Thread overview: 5+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2023-06-11 21:33 [PATCH 0/4] minor modernizations and cleanups Eric Wong
2023-06-11 21:33 ` [PATCH 1/4] avoid unnecessary #to_io calls Eric Wong
2023-06-11 21:33 ` [PATCH 2/4] my_fileno: use rb_io_check_closed for Ruby <3.1 Eric Wong
2023-06-11 21:33 ` Eric Wong [this message]
2023-06-11 21:33 ` [PATCH 4/4] test_linux_reuseport_tcp_listen_stats: skip w/o unicorn Eric Wong

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

  List information: https://yhbt.net/raindrops/

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=20230611213328.379546-4-bofh@yhbt.net \
    --to=bofh@yhbt.net \
    --cc=raindrops-public@yhbt.net \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
Code repositories for project(s) associated with this public inbox

	https://yhbt.net/raindrops.git/

This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for read-only IMAP folder(s) and NNTP newsgroup(s).