git/list[1] front-page[2] threads[3] people[4] search[5] about
 

Re: [msysGit] [PATCH/RFC 09/11] daemon: use run-command api for async serving

From
Johannes Sixt <j6t@kdbg.org>
Date
Nov 27, 2009, 20:59 UTC
Message-ID
<200911272159.38757.j6t@kdbg.org>
In-Reply-To
<1259196260-3064-10-git-send-email-kusmabite@gmail.com>
On Donnerstag, 26. November 2009, Erik Faye-Lund wrote:
Show 15 quoted lines
>  static void check_dead_children(void)
>  {
> -	int status;
> -	pid_t pid;
> -
> -	while ((pid = waitpid(-1, &status, WNOHANG)) > 0) {
> -		const char *dead = "";
> -		remove_child(pid);
> -		if (!WIFEXITED(status) || (WEXITSTATUS(status) > 0))
> -			dead = " (with error)";
> -		loginfo("[%"PRIuMAX"] Disconnected%s", (uintmax_t)pid, dead);
> -	}
> +	struct child **cradle, *blanket;
> +	for (cradle = &firstborn; (blanket = *cradle);)
> +		if (!is_async_alive(&blanket->async)) {

This would be the right place to call finish_async(). But since we cannot wait, you invented is_async_alive(). But actually we are not only interested in whether the process is alive, but also whether it completed successfully so that we can add "(with error)". Would it make sense to have a function finish_async_nowait() instead of is_async_alive() that (1) stresses the start/finish symmetry and (2) can return more than just Boolean?

> +			*cradle = blanket->next;
> +			loginfo("Disconnected\n");

Here you are losing information about the pid, which is important to have in the syslog. The \n should be dropped.

Show 13 quoted lines
> +	async.proc = async_execute;
> +	async.data = ss;
> +	async.out = incoming;
>
> -	dup2(incoming, 0);
> -	dup2(incoming, 1);
> +	if (start_async(&async))
> +		logerror("unable to fork");
> +	else
> +		add_child(&async, addr, addrlen);
>  	close(incoming);
> -
> -	exit(execute(0, addr));

In start_command(), the convention is that fds that are provided by the caller are closed by start_command() (even if there are errors). The close(incoming) that you leave here indicates that you are not using the same convention with start_async(). It would be nice to switch to the same convention.

-- Hannes
Previous: Johannes SixtNext: Erik Faye-Lund
Message 14 of 54 in “daemon-win32”
  1. 00/11 daemon-win32Erik Faye-Lund, Nov 26, 2009
  2. 01/11 mingw: add network-wrappers for daemonErik Faye-Lund, Nov 26, 2009
  3. 02/11 strbuf: add non-variadic function strbuf_vaddf()Erik Faye-Lund, Nov 26, 2009
  4. 03/11 mingw: implement syslogErik Faye-Lund, Nov 26, 2009
  5. 04/11 compat: add inet_pton and inet_ntop prototypesErik Faye-Lund, Nov 26, 2009
  6. 05/11 inet_ntop: fix a couple of old-style declsErik Faye-Lund, Nov 26, 2009
  7. 06/11 run-command: add kill_async() and is_async_alive()Erik Faye-Lund, Nov 26, 2009
  8. 07/11 run-command: support input-fdErik Faye-Lund, Nov 26, 2009
  9. 08/11 daemon: use explicit file descriptorErik Faye-Lund, Nov 26, 2009
  10. 09/11 daemon: use run-command api for async servingErik Faye-Lund, Nov 26, 2009
  11. 10/11 daemon: use full buffered mode for stderrErik Faye-Lund, Nov 26, 2009
  12. 11/11 mingw: compile git-daemonErik Faye-Lund, Nov 26, 2009
  13. Johannes SixtNov 27, 2009
  14. Johannes SixtNov 27, 2009
  15. Erik Faye-LundDec 2, 2009
  16. Johannes SixtDec 2, 2009
  17. Erik Faye-LundDec 8, 2009
  18. Johannes SixtNov 26, 2009
  19. Erik Faye-LundNov 27, 2009
  20. Erik Faye-LundNov 27, 2009
  21. Johannes SixtNov 27, 2009
  22. Johannes SixtNov 27, 2009
  23. Erik Faye-LundDec 8, 2009
  24. Johannes SixtNov 26, 2009
  25. Erik Faye-LundNov 27, 2009
  26. Johannes SixtNov 27, 2009
  27. Erik Faye-LundDec 8, 2009
  28. Johannes SixtNov 26, 2009
  29. Erik Faye-LundNov 27, 2009
  30. Johannes SixtNov 27, 2009
  31. Erik Faye-LundDec 2, 2009
  32. Johannes SixtDec 2, 2009
  33. Erik Faye-LundJan 9, 2010
  34. Erik Faye-LundJan 10, 2010
  35. Johannes SixtNov 26, 2009
  36. Erik Faye-LundNov 27, 2009
  37. Johannes SixtNov 27, 2009
  38. Erik Faye-LundDec 8, 2009
  39. Junio C HamanoNov 26, 2009
  40. Erik Faye-LundNov 26, 2009
  41. Paolo BonziniNov 26, 2009
  42. Junio C HamanoNov 26, 2009
  43. Erik Faye-LundNov 26, 2009
  44. Johannes SixtNov 27, 2009
  45. Martin StorsjöNov 26, 2009
  46. Improve the mingw getaddrinfo stub to handle more use casesMartin Storsjö, Nov 26, 2009
  47. Erik Faye-LundNov 26, 2009
  48. Martin StorsjöNov 26, 2009
  49. Erik Faye-LundDec 2, 2009
  50. Martin StorsjöDec 2, 2009
  51. Erik Faye-LundDec 2, 2009
  52. Erik Faye-LundDec 2, 2009
  53. Johannes SixtDec 2, 2009
  54. Johannes SixtNov 26, 2009

Read the whole thread, see it on lore, or plain text.

$ cat FOOTERMessages come from the public archive at lore.kernel.org/git, fetched every hour. The front page is chosen and written each morning by an AI editor and can be wrong; the threads themselves are the record. About and API. For agents: an MCP server at https://gitlist.dev/mcp, and any thread, story or person page as Markdown by adding .md to its URL (or sending Accept: text/markdown). Details in /llms.txt.