1. 08 Oct, 2021 1 commit
  2. 07 Oct, 2021 1 commit
  3. 02 Oct, 2021 2 commits
  4. 30 Sep, 2021 5 commits
    • Harald Sitter's avatar
      fix the dep5 file · 2fa7899d
      Harald Sitter authored
      it should annotate this source, not the source within the source
      2fa7899d
    • Harald Sitter's avatar
      apply mtime on smb->smb copies · 17de407c
      Harald Sitter authored
      previously this behaved wildly inconsistently with the other copy
      functions and other workers. copying on the same protocol doesn't really
      change anything, if anything it  is even more expected to not disturb
      the mtime, certainly when we were provided with a modify time, which all
      this is contingent on anyway
      17de407c
    • Harald Sitter's avatar
      stop mtime duplication · 1b464d30
      Harald Sitter authored
      applying the mtime was duplicated across multiple functions when indeed
      it is largely the same code with only the actual setting being slightly
      variable. use a template function with callback instead to share the
      sharable bits.
      
      also, since smbc_utime is used in two functions move that to its own
      helper built on top to remove even the callback duplication
      1b464d30
    • Harald Sitter's avatar
      log more flags · 82cf40ac
      Harald Sitter authored
      they can change behavior quite a bit
      82cf40ac
    • Harald Sitter's avatar
      use posix mode macros · e5c2f63e
      Harald Sitter authored
      ...to be more consistent with the rest of the code
      e5c2f63e
  5. 25 Sep, 2021 1 commit
  6. 23 Sep, 2021 1 commit
  7. 22 Sep, 2021 1 commit
    • Harald Sitter's avatar
      implement timeouts from ws-discovery · bef55f10
      Harald Sitter authored
      previously we had a global timeout in smb_browse but that was eventually
      removed to not break SMBC discovery, which is also used for listing
      directories. this left WSD without any timeout system when no WSD hosts
      are on the network because the internal timeout tracking would only
      start after the first response.
      
      instead implement the actually specified timeout mechanics of WSD.
      
      - /Probe may only be Matched within MATCH_TIMEOUT (m_probeMatchTimer =
      global timeout tracker now)
      - /Resolve may only be Matched within MATCH_TIMEOUT (encapsulated inside
      new WSDResolver)
      - PBSD may only be answered within HTTP_TIMEOUT (couldn't find a
      specified timeout so I'm guessing one would ordinarily use a default
      http timeout, since we only do a single cheap query I'm using a more
      opinionate timeout though)
      
      MATCH_TIMEOUT is derived form the DPWS spec rather than the underlying
      ws-discovery because it has more lenient timing and is generally a bit
      more assertive with requirements.
      
      BUG: 436574
      bef55f10
  8. 21 Sep, 2021 1 commit
  9. 20 Sep, 2021 1 commit
  10. 19 Sep, 2021 1 commit
  11. 16 Sep, 2021 2 commits
  12. 15 Sep, 2021 1 commit
  13. 14 Sep, 2021 1 commit
  14. 10 Sep, 2021 2 commits
  15. 09 Sep, 2021 3 commits
  16. 02 Sep, 2021 5 commits
  17. 01 Sep, 2021 2 commits
  18. 31 Aug, 2021 1 commit
  19. 25 Aug, 2021 3 commits
  20. 22 Aug, 2021 1 commit
  21. 06 Aug, 2021 1 commit
  22. 02 Aug, 2021 2 commits
  23. 28 Jul, 2021 1 commit