Hi All, Just a quick one to advise that I have a build of gcc version 4.4.0 20090414 (prerelease) (GCC) running through some tests locally - the release of GCC 4.4.0 is expected within the next week, OS/2 build will follow very shortly based on the tests I've done so far! -- -- Cheers, Paul.
On 16.04.09 07:26, Paul Smedley wrote: > Just a quick one to advise that I have a build of gcc version 4.4.0 > 20090414 (prerelease) (GCC) running through some tests locally - the > release of GCC 4.4.0 is expected within the next week, OS/2 build will > follow very shortly based on the tests I've done so far! Hi Paul, it's great that you keep updating GCC, and I'm really glad that you are working on OS/2 GCC. But I really wonder what we gain from this? For serious development even 4.3.x is useless because of the debugging incompatibility, builds still either have to be done libgcc-statically (which doesn't work in all cases and increases size by a bit) or require gcc4xx.dll. I would rather see those issues fixed in the old versions than trying to work with a new version that again has new serious bugs (as with any GCC release with minor version <~ 3). Does 4.4 have so great new features that we can benefit from? I can't find any in the list at <http://gcc.gnu.org/gcc-4.4/changes.html>. Peter.
On Fri, 17 Apr 2009 07:52:17 UTC, Peter Weilbacher <newsspam <at> weilbacher.org> wrote: > On 16.04.09 07:26, Paul Smedley wrote: > > Just a quick one to advise that I have a build of gcc version 4.4.0 > > 20090414 (prerelease) (GCC) running through some tests locally - the > > release of GCC 4.4.0 is expected within the next week, OS/2 build will > > follow very shortly based on the tests I've done so far! > > Hi Paul, > > it's great that you keep updating GCC, and I'm really glad that you are > working on OS/2 GCC. > > But I really wonder what we gain from this? For serious development even > 4.3.x is useless because of the debugging incompatibility, builds still > either have to be done libgcc-statically (which doesn't work in all cases > and increases size by a bit) or require gcc4xx.dll. I would rather see those > issues fixed in the old versions than trying to work with a new version that > again has new serious bugs (as with any GCC release with minor version <~ 3). > Does 4.4 have so great new features that we can benefit from? I can't find > any in the list at <http://gcc.gnu.org/gcc-4.4/changes.html>. > > Peter. I tried to post this previously but it didn't go through it seems. On <http://gcc.gnu.org/gcc-4.4/changes.html> it shows the following and I am curious if it affects us: Packed bit-fields of type char were not properly bit-packed on many targets prior to GCC 4.4. On these targets, the fix in GCC 4.4 causes(Continue reading)
On 18.04.09 21:58, Andy Willis wrote: > I tried to post this previously but it didn't go through it seems. On > <http://gcc.gnu.org/gcc-4.4/changes.html> it shows the following and I > am curious if it affects us: > > Packed bit-fields of type char were not properly bit-packed on many > targets prior to GCC 4.4. On these targets, the fix in GCC 4.4 causes > an ABI change. For example there is no longer a 4-bit padding between > field a and b in this structure: > > struct foo > { > char a:4; > char b:8; > } __attribute__ ((packed)); I looked through the Mozilla sources and found one instance of a similar contruct. In a place that is no part of an API, so at least for that it should be fine. P.
Hi Peter, On Fri, 17 Apr 2009 07:52:17 UTC, Peter Weilbacher <newsspam <at> weilbacher.org> wrote: > On 16.04.09 07:26, Paul Smedley wrote: > > Just a quick one to advise that I have a build of gcc version 4.4.0 > > 20090414 (prerelease) (GCC) running through some tests locally - the > > release of GCC 4.4.0 is expected within the next week, OS/2 build will > > follow very shortly based on the tests I've done so far! > > Hi Paul, > > it's great that you keep updating GCC, and I'm really glad that you are > working on OS/2 GCC. > > But I really wonder what we gain from this? For serious development even > 4.3.x is useless because of the debugging incompatibility, builds still > either have to be done libgcc-statically (which doesn't work in all cases > and increases size by a bit) or require gcc4xx.dll. I would rather see those > issues fixed in the old versions than trying to work with a new version that > again has new serious bugs (as with any GCC release with minor version <~ 3). > Does 4.4 have so great new features that we can benefit from? I can't find > any in the list at <http://gcc.gnu.org/gcc-4.4/changes.html>. Well for one thing, the gap in moving from GCC 3.3.5 to 4.3.x was quite large and required a lot of effort. By continually building each release as it comes up, the amount of changes/problems are greatly reduced.(Continue reading)
On 17.04.09 11:12, Paul Smedley wrote:
> When I get some time, I still plan to work on optlink fixes, and also
> the debugging support in emxomf. I did spend some time investigating
> the emxomf source and it did my head in - but the source for this is
> available if others want to work on it.
If I only had time, I had already looked at that... I am so busy with
other stuff (outside OS/2) by now that I can even hardly test fixes
that you come up with.
Peter.
Hi Peter, On Mon, 20 Apr 2009 22:48:07 UTC, Peter Weilbacher <newsspam <at> weilbacher.org> wrote: > On 17.04.09 11:12, Paul Smedley wrote: > > > When I get some time, I still plan to work on optlink fixes, and also > > the debugging support in emxomf. I did spend some time investigating > > the emxomf source and it did my head in - but the source for this is > > available if others want to work on it. > > If I only had time, I had already looked at that... I am so busy with > other stuff (outside OS/2) by now that I can even hardly test fixes > that you come up with. I understand 100% - my day job starts to consume more and more time - leaving me more tired at night and not willing/able to take on the really hard stuff..... -- -- Cheers, Paul.
On 04/17/09 02:12 am, Paul Smedley wrote: >> Does 4.4 have so great new features that we can benefit from? I can't find >> > any in the list at<http://gcc.gnu.org/gcc-4.4/changes.html>. > Well for one thing, the gap in moving from GCC 3.3.5 to 4.3.x was > quite large and required a lot of effort. By continually building > each release as it comes up, the amount of changes/problems are > greatly reduced. > Looking at <http://gcc.gnu.org/gcc-4.4/changes.html> I see <quote> Support for a number of older systems and recently unmaintained or untested target ports of GCC has been declared obsolete in GCC 4.4. Unless there is activity to revive them, the next release of GCC will have their sources permanently removed. The following ports for individual systems on particular architectures have been obsoleted: * Generic a.out on IA32 and m68k (i[34567]86-*-aout*, m68k-*-aout*) </quote> So this will be the last OS/2 version unless someone contacts the GCC people and let them know that we use aout and that it still works and should be maintained. Dave
On 17.04.09 16:54, Dave Yeo wrote: > The following ports for individual systems on particular architectures > have been obsoleted: > > * Generic a.out on IA32 and m68k (i[34567]86-*-aout*, m68k-*-aout*) > </quote> > > So this will be the last OS/2 version unless someone contacts the GCC > people and let them know that we use aout and that it still works and > should be maintained. I thought we used OMF format and not a.out? Ah wait, I almost always use -Zomf to be able to debug stuff. In any case, the GCC 4.4 line should carry us the remaining years of OS/2 usage.At least 3.3.5 has been reliable for 3.5 years. Peter.
Hi Peter, On Mon, 20 Apr 2009 22:43:35 UTC, Peter Weilbacher <newsspam <at> weilbacher.org> wrote: > On 17.04.09 16:54, Dave Yeo wrote: > > > The following ports for individual systems on particular architectures > > have been obsoleted: > > > > * Generic a.out on IA32 and m68k (i[34567]86-*-aout*, m68k-*-aout*) > > </quote> > > > > So this will be the last OS/2 version unless someone contacts the GCC > > people and let them know that we use aout and that it still works and > > should be maintained. > > I thought we used OMF format and not a.out? Ah wait, I almost always > use -Zomf to be able to debug stuff. We always compile in a.out format - if your specify -Zomf as a CFLAG or LDFLAG emxomf is used to convert the a.out format to OMF format > In any case, the GCC 4.4 line should carry us the remaining years of > OS/2 usage.(Continue reading)
RSS Feed