python / cpython

The Python programming language
https://www.python.org
Other
63.08k stars 30.21k forks source link

diffs for Windows CE - Include/Opcode.h #35781

Closed a77a3d74-e722-44ae-97d2-39445fc24e84 closed 22 years ago

a77a3d74-e722-44ae-97d2-39445fc24e84 commented 22 years ago
BPO 494783
Nosy @gvanrossum, @loewis
Files
  • opcode.diff: Include/Opcode.h context diff for Wince
  • opcode.diff: opcode.diff
  • compile.c.diff: compile.c.diff
  • ceval.diff: ceval.diff
  • Note: these values reflect the state of the issue at the time it was migrated and might not reflect the current state.

    Show more details

    GitHub fields: ```python assignee = None closed_at = created_at = labels = ['interpreter-core'] title = 'diffs for Windows CE - Include/Opcode.h' updated_at = user = 'https://bugs.python.org/bkc' ``` bugs.python.org fields: ```python activity = actor = 'loewis' assignee = 'none' closed = True closed_date = None closer = None components = ['Interpreter Core'] creation = creator = 'bkc' dependencies = [] files = ['3859', '3860', '3861', '3862'] hgrepos = [] issue_num = 494783 keywords = ['patch'] message_count = 12.0 messages = ['38515', '38516', '38517', '38518', '38519', '38520', '38521', '38522', '38523', '38524', '38525', '38526'] nosy_count = 3.0 nosy_names = ['gvanrossum', 'loewis', 'bkc'] pr_nums = [] priority = 'normal' resolution = 'accepted' stage = None status = 'closed' superseder = None type = None url = 'https://bugs.python.org/issue494783' versions = [] ```

    a77a3d74-e722-44ae-97d2-39445fc24e84 commented 22 years ago

    Attached is a diff for Include/opcode.h to allow compilation on MS Wince using EVT 3.0

    I have quite a lot of little diffs like this that I would like to submit, but before generating and posting all of them, I'm sending in just this one to be sure I've done this correctly.

    My original edits were made against 2.2 alpha 1, so I am re-updating my local src tree and redoing all the diffs.

    I also will have some diffs for Novell NetWare in the future.

    I realize you may not be able to apply these diffs anytime soon, but could you let me know if this meets your format requirements soon so I can continue to submit diffs.

    Thanks

    gvanrossum commented 22 years ago

    Logged In: YES user_id=6380

    That looks like a fine context diff to me.

    Please be sure to use the current CVS if you can!

    a77a3d74-e722-44ae-97d2-39445fc24e84 commented 22 years ago

    Logged In: YES user_id=4631

    I used the win CVS "update" command on the entire source tree yesterday. I have not specified a particular tag. Are you saying that opcode.h rev 2.37 is not current, or is "please use the current cvs" a standard cautionary note?

    Sorry to be anal, I want this to be seamless for you so I feel I have botched it up already.

    gvanrossum commented 22 years ago

    Logged In: YES user_id=6380

    Standard cautionary note. Sorry.

    61337411-43fc-4a9c-b8d5-4060aede66d0 commented 22 years ago

    Logged In: YES user_id=21627

    The patch looks fine to me. I wonder whether Python should stop using these constant names, though, and replace them with, say, a PyCmp_ prefix, throughout. If there is a backwards compatibility concern (which I wouldn't expect), it should be possible to add #defines to the old names when there is no conflict.

    Would you be willing to rewrite this patch to do so?

    a77a3d74-e722-44ae-97d2-39445fc24e84 commented 22 years ago

    Logged In: YES user_id=4631

    Yes, I will change the enum and all uses of it that I can find to use a unique prefix as you suggested.

    a77a3d74-e722-44ae-97d2-39445fc24e84 commented 22 years ago

    Logged In: YES user_id=4631

    I have made the suggested changes to the enum, and the necessary changes to compile.c and ceval.c

    Should I append those three diffs to this patch entry, or create a new entry, one for each of those diffs?

    compile.c has some other unrelated changes, and some of those changes are dependent on diffs to another include.

    IE, rather than #ifdef'ing all references to errno, I know that NetWare doesn't have errno either. So I've created macros GetErrno(), ClearErrno() and SetErrno(). Hope that's going to work out.

    Also, lots of modules have statements like : goto finally; which Metrowerks and EVT don't like because "finally" is a reserved word, even when compiling .c modules. So I've had to change all of those as well.

    When I get regrtests all working, and recompile my code on Win32 and Linux, then I'll post all my diffs.

    a77a3d74-e722-44ae-97d2-39445fc24e84 commented 22 years ago

    Logged In: YES user_id=4631

    I have made the suggested changes to the enum, and the necessary changes to compile.c and ceval.c

    Should I append those three diffs to this patch entry, or create a new entry, one for each of those diffs?

    compile.c has some other unrelated changes, and some of those changes are dependent on diffs to another include.

    IE, rather than #ifdef'ing all references to errno, I know that NetWare doesn't have errno either. So I've created macros GetErrno(), ClearErrno() and SetErrno(). Hope that's going to work out.

    Also, lots of modules have statements like : goto finally; which Metrowerks and EVT don't like because "finally" is a reserved word, even when compiling .c modules. So I've had to change all of those as well.

    When I get regrtests all working, and recompile my code on Win32 and Linux, then I'll post all my diffs.

    61337411-43fc-4a9c-b8d5-4060aede66d0 commented 22 years ago

    Logged In: YES user_id=21627

    Attaching the opcode changes to this report is fine, also including any other renaming problems into the same patch is fine.

    I'd appreciate if you could separate any errno patches you may have, since I'd expect some discussion on these, which shouldn't stop integration of the renamings. To do this, you could try to edit the universal or context diffs (whatever you are more familiar with) by hand. Notice that this works only as long as you are deleting entire hunks; don't try to modify a hunk (this will normally mess up the line numbers).

    If overlapping changes are in a single hunk, you may include them in the patch as well - I'll have to unedit them out, then (by applying the patch to a clean copy, and manually taking out the unrelated changes).

    You may wonder about the insistence on separate patches, but it is really a prerequisite for understanding the rationale of a change in the coming years.

    a77a3d74-e722-44ae-97d2-39445fc24e84 commented 22 years ago

    Logged In: YES user_id=4631

    Here are three updated diff's for ceval.c, compile.c and opcode.h I have removed the errno-macro components of the diff. When I go to submit the errno-diffs, will I need to remove the diffs that are included here? (lots of editing!)

    a77a3d74-e722-44ae-97d2-39445fc24e84 commented 22 years ago

    Logged In: YES user_id=4631

    The opcode.diff file with the description "Include/Opcode.h context diff for Wince" should be deleted. I don't have rights to delete it.

    61337411-43fc-4a9c-b8d5-4060aede66d0 commented 22 years ago

    Logged In: YES user_id=21627

    Thanks for the patch, committed as opcode.h 2.38 ceval.c 2.303 compile.c 2.235 ACKS 1.153 NEWS 1.344

    Please update your code base; this should solve the problem of sorting out the additional changes.