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

Re: [PATCH v3 13/23] log_ref_setup(): pass the open file descriptor back to the caller

From
Jeff King <peff@peff.net>
Date
Dec 31, 2016, 06:32 UTC
Message-ID
<20161231063211.tqsiafg3iahcuotz@sigill.intra.peff.net>
In-Reply-To
<ef2355e9d5ccaa53928c821530bae59f2b118013.1483153436.git.mhagger@alum.mit.edu>
On Sat, Dec 31, 2016 at 04:12:53AM +0100, Michael Haggerty wrote:
Show 5 quoted lines
> This function will most often be called by log_ref_write_1(), which
> wants to append to the reflog file. In that case, it is silly to close
> the file only for the caller to reopen it immediately. So, in the case
> that the file was opened, pass the open file descriptor back to the
> caller.
Sounds like a much saner interface.
Show 18 quoted lines
>  /*
> - * Create a reflog for a ref.  If force_create = 0, the reflog will
> - * only be created for certain refs (those for which
> - * should_autocreate_reflog returns non-zero.  Otherwise, create it
> - * regardless of the ref name.  Fill in *err and return -1 on failure.
> + * Create a reflog for a ref. Store its path to *logfile. If
> + * force_create = 0, only create the reflog for certain refs (those
> + * for which should_autocreate_reflog returns non-zero). Otherwise,
> + * create it regardless of the reference name. If the logfile already
> + * existed or was created, return 0 and set *logfd to the file
> + * descriptor opened for appending to the file. If no logfile exists
> + * and we decided not to create one, return 0 and set *logfd to -1. On
> + * failure, fill in *err, set *logfd to -1, and return -1.
>   */
> -static int log_ref_setup(const char *refname, struct strbuf *logfile, struct strbuf *err, int force_create)
> +static int log_ref_setup(const char *refname,
> +			 struct strbuf *logfile, int *logfd,
> +			 struct strbuf *err, int force_create)

The return value is always "0" or "-1". It seems like it would be simpler to just return the descriptor instead of 0.

I guess that makes it hard to identify the case when we chose not to create a descriptor. I wonder if more "normal" semantics would be:

  1. ret >= 0: file existed or was created, and ret is the descriptor
  2. ret < 0, err is empty: we chose not to create
  3. ret < 0, err is non-empty: a real error

I dunno. This may just be bikeshedding, and I can live with it either way (especially because you documented it!).

-Peff
Previous: Michael HaggertyNext: Michael Haggerty
Message 12 of 55 in “Delete directories left empty after ref deletion”
  1. 00/23 Delete directories left empty after ref deletionMichael Haggerty, Dec 31, 2016
  2. 01/23 files_rename_ref(): tidy up whitespaceMichael Haggerty, Dec 31, 2016
  3. 02/23 t5505: use "for-each-ref" to test for the non-existence of referencesMichael Haggerty, Dec 31, 2016
  4. 03/23 safe_create_leading_directories_const(): preserve errnoMichael Haggerty, Dec 31, 2016
  5. 05/23 raceproof_create_file(): new functionMichael Haggerty, Dec 31, 2016
  6. Jeff KingDec 31, 2016
  7. Michael HaggertyDec 31, 2016
  8. Junio C HamanoJan 1, 2017
  9. 06/23 lock_ref_sha1_basic(): inline constantMichael Haggerty, Dec 31, 2016
  10. 07/23 lock_ref_sha1_basic(): use raceproof_create_file()Michael Haggerty, Dec 31, 2016
  11. 13/23 log_ref_setup(): pass the open file descriptor back to the callerMichael Haggerty, Dec 31, 2016
  12. Jeff KingDec 31, 2016
  13. Michael HaggertyDec 31, 2016
  14. Jeff KingDec 31, 2016
  15. Junio C HamanoJan 1, 2017
  16. 12/23 log_ref_setup(): improve robustness against racesMichael Haggerty, Dec 31, 2016
  17. 11/23 log_ref_setup(): separate code for create vs non-createMichael Haggerty, Dec 31, 2016
  18. Jeff KingDec 31, 2016
  19. Michael HaggertyDec 31, 2016
  20. Junio C HamanoJan 1, 2017
  21. Michael HaggertyJan 1, 2017
  22. 15/23 log_ref_setup(): manage the name of the reflog file internallyMichael Haggerty, Dec 31, 2016
  23. 16/23 log_ref_write_1(): inline functionMichael Haggerty, Dec 31, 2016
  24. 19/23 try_remove_empty_parents(): rename parameter "name" -> "refname"Michael Haggerty, Dec 31, 2016
  25. 20/23 try_remove_empty_parents(): don't trash argument contentsMichael Haggerty, Dec 31, 2016
  26. Jeff KingDec 31, 2016
  27. Michael HaggertyJan 2, 2017
  28. Jeff KingJan 2, 2017
  29. 22/23 try_remove_empty_parents(): teach to remove parents of reflogs, tooMichael Haggerty, Dec 31, 2016
  30. 21/23 try_remove_empty_parents(): don't accommodate consecutive slashesMichael Haggerty, Dec 31, 2016
  31. Junio C HamanoJan 1, 2017
  32. Jeff KingJan 1, 2017
  33. Michael HaggertyJan 2, 2017
  34. Jeff KingJan 2, 2017
  35. 23/23 files_transaction_commit(): clean up empty directoriesMichael Haggerty, Dec 31, 2016
  36. 04/23 safe_create_leading_directories(): set errno on SCLD_EXISTSMichael Haggerty, Dec 31, 2016
  37. 08/23 rename_tmp_log(): use raceproof_create_file()Michael Haggerty, Dec 31, 2016
  38. 14/23 log_ref_write_1(): don't depend on logfile argumentMichael Haggerty, Dec 31, 2016
  39. Jeff KingDec 31, 2016
  40. Michael HaggertyDec 31, 2016
  41. 17/23 delete_ref_loose(): derive loose reference path from lockMichael Haggerty, Dec 31, 2016
  42. 18/23 delete_ref_loose(): inline functionMichael Haggerty, Dec 31, 2016
  43. 09/23 rename_tmp_log(): improve error reportingMichael Haggerty, Dec 31, 2016
  44. 10/23 log_ref_write(): inline functionMichael Haggerty, Dec 31, 2016
  45. Junio C HamanoJan 1, 2017
  46. Michael HaggertyJan 1, 2017
  47. Jeff KingDec 31, 2016
  48. Junio C HamanoJan 1, 2017
  49. Jacob KellerJan 1, 2017
  50. Jacob KellerJan 1, 2017
  51. Philip OakleyJan 1, 2017
  52. Jacob KellerJan 1, 2017
  53. Jeff KingJan 2, 2017
  54. Michael HaggertyJan 2, 2017
  55. Jacob KellerJan 2, 2017

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.