QueueGameCommand optimization

[armoha]

• [ ] Optimize for multiple calls in single trigger cycle.

  • Currently every QueueGameCommand call reads command queue length and max buffer size.

  https://github.com/armoha/eudplib/blob/ba9c840a143fe8d764caebcca3b4adfb77a5188c/eudplib/eudlib/qgcf/qgc.py#L49-53

  Potential improvement:

  MAX_BUFFER = 0x57F0D8
  CMDQ_LEN = 0x654AA0
  cmdqlen = c.Forward()
  if cs.EUDIfNot()(c.Memory(CMDQ_LEN, c.Exactly, cmdqlen)):
      cmdqlen << f_dwread_epd(ut.EPD(CMDQ_LEN))
  cs.EUDEndIf()
  if cs.EUDIf()(c.Memory(MAX_BUFFER, c.AtLeast, cmdqlen + size)):
      f_memcpy(0x654880 + cmdqlen, data, size)
      cmdqlen += size
      c.SetVariables(ut.EPD(0x654AA0), cmdqlen)
  cs.EUDEndIf()

[armoha]

Almost fixed by https://github.com/armoha/eudplib/commit/1398bf86f5c37ce2375eb8db85422c560acaacfe and https://github.com/armoha/eudplib/commit/2051c778ec48156388a8fb0cf3b3cd426c5afcb7. One thing left to solve is preserving offset state 0x654880 + qgc.cmdqlen of EUDByteWriter for consecutive QGC calls.