Re: [libposix-development] My personal development branch
Status: Pre-Alpha
Brought to you by:
hdante
Menu
▾
▴
Re: [libposix-development] My personal development branch
|
From: Henrique A. <hd...@gm...> - 2009-06-14 18:08:34
|
I don't think the code rules are overly stringent. The method to deal with real world implementations was described in the last post (so I don't need to repeat it here). I don't think organizing things is ridiculous either. Actually, I'm very proud of doing that way. See you, 2009/6/14 Andreas Krennmair <akr...@go...>: > Linux is not a standard, but Linux is real! If you want to support > certain system calls on Linux, you must adhere to Linux's data > structures. Or are you trying to suggest that I should write a wrapper > around such things that map the Linux data structures to data > structures that adhere to POSIX.1-2008 and POSIX.1-2008 only? That > would be just outright crazy. > > Anyway, your tight restrictions on the code and all your fighting > about the confrontation with real-world conditions really took my last > motivation to contribute my code. Apparently, working on the code > itself for this project is fun, and I will keep developing in my own > development branch, but don't expect me to report any improvements or > changes that I made here. > > Good luck to you and all the other contributing C developers, low > level developers, testers, web designers, web developers, web > editors/bloggers/"PR guys", build system developers/port maintainers, > performance developers, roadmap managers and code reviewers. With more > stringent code conformity rules and tighter project roles, you will > surely make it! > > Best regards, > Andreas > > On Sat, Jun 13, 2009 at 11:14 PM, Henrique Almeida<hd...@gm...> wrote: >> No, that won't do. Submitted code must follow the standard (or must >> go to system specific directories). It doesn't make sense for the >> project otherwise. >> >> - Code that is contributed must be written for the standard. Linux is >> not a standard. (for example, sigval_t is not defined in the standard, >> sigval is, __band_t is not defined either). >> - System specific code that must go to the system directory. They >> always need a reference that proves that they're necessary. >> - Code that is copied from other libraries must have original >> author's notice, so that it can be tagged as derived work (and the >> license must allow that). >> >> 2009/6/13 Andreas Krennmair <akr...@go...>: >>> Henrique, >>> >>> I'm not pulling the struct definition out of my a**, Linux is the way >>> it is (especially its kernel data structures), and there's nothing >>> that can be changed about it from my side. If this is unacceptable, >>> well, then I can't help you. Take the code or leave it, it's all your >>> decision. Having to vindicate every single line of code that I >>> contribute isn't exactly something that makes me eager to contribute >>> more code. >>> >>> Best regards, >>> Andreas >>> >>> On Sat, Jun 13, 2009 at 5:21 AM, Henrique Almeida<hd...@gm...> wrote: >>>> Comments for commit: >>>> >>>> 1e39bb5359952c6c6e28cc74f6651c98c8d86905 added auto-generated system >>>> calls (including auto-generator script). >>>> >>>> signal.h: >>>> - the "include" files can't contain conditional code (_WORDSIZE), >>>> move signal.h to system specific directories (or add an #include >>>> <arch_signal.h>) >>>> stat.h: >>>> - __pad0, __st_ino, etc. where did those come from ? >>>> - can't have __attribute__ extension in "include" directory (where >>>> did those come from ?) >>>> - sigval_t this doesn't exist in the standard >>>> >>>> Are you copying this from an older implementation ? There are too many >>>> unexpected features in this patch. I need to understand what's this >>>> all about before I continue reading the patches. I can't spend time >>>> verifying that non standard code is being submitted, this has to be >>>> done by the contributor. >>>> >>>> 2009/6/12 Henrique Almeida <hd...@gm...>: >>>>> Comments for commit: >>>>> >>>>> 9a0a36a4cbfdea5f8666cf52747fb7dc317258ab fixed i386 build. >>>>> >>>>> - uncomment #ifdef in syscall_nums.h >>>>> - vsyscall address in being overwritten in line 108, movl 28(%ebp), %edi >>>>> >>>>> 2009/6/9 Henrique Almeida <hd...@gm...>: >>>>>> Broken branch updated up to 2009-06-05 >>>>>> >>>>>> 2009/6/9 Henrique Almeida <hd...@gm...>: >>>>>>> Comments for commit: >>>>>>> >>>>>>> 3ea871f2ae8e772127d0d2434f027ad4d2586678 implemented strrchr() >>>>>>> including unit tests. >>>>>>> >>>>>>> on strrchr.c, '\0' must be considered part of the string, fix both >>>>>>> the implementation and test >>>>>>> >>>>>>> 2009/6/9 Henrique Almeida <hd...@gm...>: >>>>>>>> Comments for commit: >>>>>>>> >>>>>>>> d808005b6f54ae16bdc5e18d4604afe6c6d7949d added unit tests for already >>>>>>>> implemented str* functions. >>>>>>>> >>>>>>>> - remove test prefix from file names >>>>>>>> - test.strchr1.c third test (TEST_ASSERT(strchr(buf, '\0')==NULL)) is >>>>>>>> incorrect, fix to TEST_ASSERT(strchr(buf, '\0')==buf+13 >>>>>>>> - add test coverage to strcmp: range 0, UCHAR_MAX >>>>>>>> - add test coverage to strcpy: out of bound writes (both sides) >>>>>>>> >>>>>>>> >>>>>>>> 2009/6/9 Henrique Almeida <hd...@gm...>: >>>>>>>>> Comments for commit: >>>>>>>>> >>>>>>>>> 1d146af4f761c6c43dbc911fca0b8fbe7d7f487d added mmap and munmap. >>>>>>>>> >>>>>>>>> - sys/mman.h >>>>>>>>> - the PROT_* macros are system dependent (they depend on the kernel used). >>>>>>>>> mman.h should be moved to the system directory. Also, the reference for >>>>>>>>> the values for linux are missing. >>>>>>>>> >>>>>>>>> >>>>>>>>> 2009/6/9 Henrique Almeida <hd...@gm...>: >>>>>>>>>> >>>>>>>>>> I have done a git pull here and there are individual commits. I think I >>>>>>>>>> haven't noticed them earlier. I'll begin to read each of them. >>>>>>>>>> >>>>>>>>>> 2009/6/8 Henrique Almeida <hd...@gm...> >>>>>>>>>>> >>>>>>>>>>> The problem was that the merge was condensed into a single commit (or >>>>>>>>>>> maybe I didn't notice the other commits in the log). I can try again later, >>>>>>>>>>> but at the moment there are some merge issues to resolve. >>>>>>>>>>> >>>>>>>>>>> 2009/6/8 Andreas Krennmair <akr...@go...> >>>>>>>>>>>> >>>>>>>>>>>> If there are differences, you always need an extra merge commit, as far >>>>>>>>>>>> as I understand. >>>>>>>>>>>> >>>>>>>>>>>> On Sat, Jun 6, 2009 at 10:40 PM, Henrique Almeida <hd...@gm...> >>>>>>>>>>>> wrote: >>>>>>>>>>>>> >>>>>>>>>>>>> With git pull, I'm not being able to merge each commit individually in >>>>>>>>>>>>> my branch: >>>>>>>>>>>>> >>>>>>>>>>>>> commit 3c75fe8dccc28d9011bb8f5a548db0778eae54cd >>>>>>>>>>>>> Merge: 1faac1a... 05b56ea... >>>>>>>>>>>>> Author: Henrique <hd...@us...> >>>>>>>>>>>>> Date: Sat Jun 6 17:29:37 2009 -0300 >>>>>>>>>>>>> >>>>>>>>>>>>> Merge git://github.com/akrennmair/libposix into andreas >>>>>>>>>>>>> >>>>>>>>>>>>> * git://github.com/akrennmair/libposix: >>>>>>>>>>>>> implemented abs(), labs() and llabs() including unit tests. >>>>>>>>>>>>> implemented strrchr() including unit tests. >>>>>>>>>>>>> fixed build dependencies for tests. >>>>>>>>>>>>> added unit tests for already implemented str* functions. >>>>>>>>>>>>> added mmap and munmap. >>>>>>>>>>>>> added simple unit tests for already implemented mem* functions. >>>>>>>>>>>>> >>>>>>>>>>>>> Is there a way to fix this ? >>>>>>>>>>>>> >>>>>>>>>>>>> 2009/6/5 Henrique Almeida <hd...@gm...> >>>>>>>>>>>>>> >>>>>>>>>>>>>> Ok, good idea, git branches rock ! >>>>>>>>>>>>>> >>>>>>>>>>>>>> 2009/6/5 Andreas Krennmair <akr...@go...> >>>>>>>>>>>>>>> >>>>>>>>>>>>>>> Hello everybody, >>>>>>>>>>>>>>> >>>>>>>>>>>>>>> As the "broken" branch didn't work out as expected, I found a better >>>>>>>>>>>>>>> way that both satisfies me in its abilities and also allows other developers >>>>>>>>>>>>>>> to easily see and merge my changes. I have a fork of libposix uploaded on >>>>>>>>>>>>>>> GitHub: http://github.com/akrennmair/libposix/tree/master >>>>>>>>>>>>>>> In this fork, I will regularly keep track of changes in the official >>>>>>>>>>>>>>> libposix git repository. But it's also simple to merge my commits into the >>>>>>>>>>>>>>> official libposix repository by executing >>>>>>>>>>>>>>> git pull git://github.com/akrennmair/libposix.git master >>>>>>>>>>>>>>> in a clone of the official libposix git repository. Anyway, GitHub is >>>>>>>>>>>>>>> the place to see and track my changes. Of course, this shall NOT be >>>>>>>>>>>>>>> interpreted as a way to get independent from the official project, it is >>>>>>>>>>>>>>> merely a way that helps me work in a more flexible fashion on the things >>>>>>>>>>>>>>> that I want to work on, plus it makes my changes more visible and also >>>>>>>>>>>>>>> easier to reintegrate with the official development line. >>>>>>>>>>>>>>> >>>>>>>>>>>>>>> Nevertheless, I will continue posting short notifications about the >>>>>>>>>>>>>>> things I currently work on to keep everyone informed. So far, I've worked on >>>>>>>>>>>>>>> some simple functions including unit tests, and I hope to present a simple >>>>>>>>>>>>>>> implementation of malloc(), free(), realloc() and calloc() within a day or >>>>>>>>>>>>>>> two. >>>>>>>>>>>>>>> >>>>>>>>>>>>>>> Regards, >>>>>>>>>>>>>>> Andreas >>>>>>>>>>>>>>> >>>>>>>>>>>>>>> >>>>>>>>>>>>>>> ------------------------------------------------------------------------------ >>>>>>>>>>>>>>> OpenSolaris 2009.06 is a cutting edge operating system for enterprises >>>>>>>>>>>>>>> looking to deploy the next generation of Solaris that includes the >>>>>>>>>>>>>>> latest >>>>>>>>>>>>>>> innovations from Sun and the OpenSource community. Download a copy and >>>>>>>>>>>>>>> enjoy capabilities such as Networking, Storage and Virtualization. >>>>>>>>>>>>>>> Go to: http://p.sf.net/sfu/opensolaris-get >>>>>>>>>>>>>>> _______________________________________________ >>>>>>>>>>>>>>> Libposix-development mailing list >>>>>>>>>>>>>>> Lib...@li... >>>>>>>>>>>>>>> https://lists.sourceforge.net/lists/listinfo/libposix-development >>>>>>>>>>>>>>> >>>>>>>>>>>>>> >>>>>>>>>>>>>> >>>>>>>>>>>>>> >>>>>>>>>>>>>> -- >>>>>>>>>>>>>> Henrique Dante de Almeida >>>>>>>>>>>>>> hd...@gm... >>>>>>>>>>>>> >>>>>>>>>>>>> >>>>>>>>>>>>> >>>>>>>>>>>>> -- >>>>>>>>>>>>> Henrique Dante de Almeida >>>>>>>>>>>>> hd...@gm... >>>>>>>>>>>>> >>>>>>>>>>>>> >>>>>>>>>>>>> ------------------------------------------------------------------------------ >>>>>>>>>>>>> OpenSolaris 2009.06 is a cutting edge operating system for enterprises >>>>>>>>>>>>> looking to deploy the next generation of Solaris that includes the >>>>>>>>>>>>> latest >>>>>>>>>>>>> innovations from Sun and the OpenSource community. Download a copy and >>>>>>>>>>>>> enjoy capabilities such as Networking, Storage and Virtualization. >>>>>>>>>>>>> Go to: http://p.sf.net/sfu/opensolaris-get >>>>>>>>>>>>> _______________________________________________ >>>>>>>>>>>>> Libposix-development mailing list >>>>>>>>>>>>> Lib...@li... >>>>>>>>>>>>> https://lists.sourceforge.net/lists/listinfo/libposix-development >>>>>>>>>>>>> >>>>>>>>>>>> >>>>>>>>>>>> >>>>>>>>>>>> >>>>>>>>>>>> ------------------------------------------------------------------------------ >>>>>>>>>>>> OpenSolaris 2009.06 is a cutting edge operating system for enterprises >>>>>>>>>>>> looking to deploy the next generation of Solaris that includes the latest >>>>>>>>>>>> innovations from Sun and the OpenSource community. Download a copy and >>>>>>>>>>>> enjoy capabilities such as Networking, Storage and Virtualization. >>>>>>>>>>>> Go to: http://p.sf.net/sfu/opensolaris-get >>>>>>>>>>>> _______________________________________________ >>>>>>>>>>>> Libposix-development mailing list >>>>>>>>>>>> Lib...@li... >>>>>>>>>>>> https://lists.sourceforge.net/lists/listinfo/libposix-development >>>>>>>>>>>> >>>>>>>>>>> >>>>>>>>>>> >>>>>>>>>>> >>>>>>>>>>> -- >>>>>>>>>>> Henrique Dante de Almeida >>>>>>>>>>> hd...@gm... >>>>>>>>>> >>>>>>>>>> >>>>>>>>>> >>>>>>>>>> -- >>>>>>>>>> Henrique Dante de Almeida >>>>>>>>>> hd...@gm... >>>>>>>>>> >>>>>>>>> >>>>>>>>> >>>>>>>>> >>>>>>>>> -- >>>>>>>>> Henrique Dante de Almeida >>>>>>>>> hd...@gm... >>>>>>>>> >>>>>>>> >>>>>>>> >>>>>>>> >>>>>>>> -- >>>>>>>> Henrique Dante de Almeida >>>>>>>> hd...@gm... >>>>>>>> >>>>>>> >>>>>>> >>>>>>> >>>>>>> -- >>>>>>> Henrique Dante de Almeida >>>>>>> hd...@gm... >>>>>>> >>>>>> >>>>>> >>>>>> >>>>>> -- >>>>>> Henrique Dante de Almeida >>>>>> hd...@gm... >>>>>> >>>>> >>>>> >>>>> >>>>> -- >>>>> Henrique Dante de Almeida >>>>> hd...@gm... >>>>> >>>> >>>> >>>> >>>> -- >>>> Henrique Dante de Almeida >>>> hd...@gm... >>>> >>>> ------------------------------------------------------------------------------ >>>> Crystal Reports - New Free Runtime and 30 Day Trial >>>> Check out the new simplified licensing option that enables unlimited >>>> royalty-free distribution of the report engine for externally facing >>>> server and web deployment. >>>> http://p.sf.net/sfu/businessobjects >>>> _______________________________________________ >>>> Libposix-development mailing list >>>> Lib...@li... >>>> https://lists.sourceforge.net/lists/listinfo/libposix-development >>>> >>> >>> ------------------------------------------------------------------------------ >>> Crystal Reports - New Free Runtime and 30 Day Trial >>> Check out the new simplified licensing option that enables unlimited >>> royalty-free distribution of the report engine for externally facing >>> server and web deployment. >>> http://p.sf.net/sfu/businessobjects >>> _______________________________________________ >>> Libposix-development mailing list >>> Lib...@li... >>> https://lists.sourceforge.net/lists/listinfo/libposix-development >>> >> >> >> >> -- >> Henrique Dante de Almeida >> hd...@gm... >> >> ------------------------------------------------------------------------------ >> Crystal Reports - New Free Runtime and 30 Day Trial >> Check out the new simplified licensing option that enables unlimited >> royalty-free distribution of the report engine for externally facing >> server and web deployment. >> http://p.sf.net/sfu/businessobjects >> _______________________________________________ >> Libposix-development mailing list >> Lib...@li... >> https://lists.sourceforge.net/lists/listinfo/libposix-development >> > > ------------------------------------------------------------------------------ > Crystal Reports - New Free Runtime and 30 Day Trial > Check out the new simplified licensing option that enables unlimited > royalty-free distribution of the report engine for externally facing > server and web deployment. > http://p.sf.net/sfu/businessobjects > _______________________________________________ > Libposix-development mailing list > Lib...@li... > https://lists.sourceforge.net/lists/listinfo/libposix-development > -- Henrique Dante de Almeida hd...@gm... |
