[Buildroot] gcc 4.2.2 patch

[Buildroot] gcc 4.2.2 patch

[Nigel Kukard]

Is this patch ok to bump gcc to 4.2.2?

I grep -r'd the buildroot tree and replaced each occurrence manually and
removed 4.2.0.

-N
-------------- next part --------------
A non-text attachment was scrubbed...
Name: gcc-4.2.2.patch
Type: text/x-patch
Size: 18843 bytes
Desc: not available
Url : http://busybox.net/lists/buildroot/attachments/20080328/5c11163b/attachment.bin 
-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: application/pgp-signature
Size: 189 bytes
Desc: This is a digitally signed message part
Url : http://busybox.net/lists/buildroot/attachments/20080328/5c11163b/attachment.pgp 

[Buildroot] gcc 4.2.2 patch

[Bernhard Fischer]

On Fri, Mar 28, 2008 at 08:22:02PM +0000, Nigel Kukard wrote:
>Is this patch ok to bump gcc to 4.2.2?
>
>I grep -r'd the buildroot tree and replaced each occurrence manually and
>removed 4.2.0.
>
>-N

>Index: package/lvm2/lvm2.mk
>===================================================================
>--- package/lvm2/lvm2.mk	(revision 21539)
>+++ package/lvm2/lvm2.mk	(working copy)
>@@ -64,7 +64,7 @@
> $(LVM2_TARGET_SBINS): $(LVM2_DIR)/.configured
> 	$(MAKE) CC=$(TARGET_CC) -C $(LVM2_DIR) DESTDIR=$(STAGING_DIR)
> 	$(MAKE) CC=$(TARGET_CC) -C $(LVM2_DIR) DESTDIR=$(STAGING_DIR) install
>-	for binary in $(LVM2_SBIN); do echo $$binary; cp -a $(STAGING_DIR)/sbin/$$binary $(TARGET_DIR)/sbin; done
>+	for binary in $(LVM2_SBIN); do echo $$binary; install -m0755 -p $(STAGING_DIR)/sbin/$$binary $(TARGET_DIR)/sbin; done

close.
Correct would have been to use $(INSTALL). Also the for loop doesn't
make too much sense, fwiw.

Something like this should do everything that is needed and is not
bloated:
	$(INSTALL) -D -m0755 $(patsubst %,$(STAGING_DIR)/sbin/%,$(LVM2_SBIN)) $(TARGET_DIR)/sbin/

Remember that this is *make* and not your favourite shell..

[Buildroot] gcc 4.2.2 patch

[Peter Korsgaard]

>>>>> "Bernhard" == Bernhard Fischer <rep.dot.nop@gmail.com> writes:

 Bernhard> On Fri, Mar 28, 2008 at 08:22:02PM +0000, Nigel Kukard wrote:
 >> Is this patch ok to bump gcc to 4.2.2?
 >> 
 >> I grep -r'd the buildroot tree and replaced each occurrence manually and
 >> removed 4.2.0.
 >> 
 >> -N

 >> Index: package/lvm2/lvm2.mk
 >> ===================================================================
 >> --- package/lvm2/lvm2.mk	(revision 21539)
 >> +++ package/lvm2/lvm2.mk	(working copy)
 >> @@ -64,7 +64,7 @@
 >> $(LVM2_TARGET_SBINS): $(LVM2_DIR)/.configured
 >> $(MAKE) CC=$(TARGET_CC) -C $(LVM2_DIR) DESTDIR=$(STAGING_DIR)
 >> $(MAKE) CC=$(TARGET_CC) -C $(LVM2_DIR) DESTDIR=$(STAGING_DIR) install
 >> -	for binary in $(LVM2_SBIN); do echo $$binary; cp -a $(STAGING_DIR)/sbin/$$binary $(TARGET_DIR)/sbin; done
 >> +	for binary in $(LVM2_SBIN); do echo $$binary; install -m0755 -p $(STAGING_DIR)/sbin/$$binary $(TARGET_DIR)/sbin; done

 Bernhard> close.
 Bernhard> Correct would have been to use $(INSTALL). Also the for
 Bernhard> loop doesn't make too much sense, fwiw.

What does this hunk have to do with $SUBJ ?

-- 
Bye, Peter Korsgaard

[Buildroot] gcc 4.2.2 patch

[Bernhard Fischer]

On Fri, Mar 28, 2008 at 09:59:44PM +0100, Peter Korsgaard wrote:
>>>>>> "Bernhard" == Bernhard Fischer <rep.dot.nop@gmail.com> writes:
>
> Bernhard> On Fri, Mar 28, 2008 at 08:22:02PM +0000, Nigel Kukard wrote:
> >> Is this patch ok to bump gcc to 4.2.2?
> >> 
> >> I grep -r'd the buildroot tree and replaced each occurrence manually and
> >> removed 4.2.0.
> >> 
> >> -N
>
> >> Index: package/lvm2/lvm2.mk
> >> ===================================================================
> >> --- package/lvm2/lvm2.mk	(revision 21539)
> >> +++ package/lvm2/lvm2.mk	(working copy)
> >> @@ -64,7 +64,7 @@
> >> $(LVM2_TARGET_SBINS): $(LVM2_DIR)/.configured
> >> $(MAKE) CC=$(TARGET_CC) -C $(LVM2_DIR) DESTDIR=$(STAGING_DIR)
> >> $(MAKE) CC=$(TARGET_CC) -C $(LVM2_DIR) DESTDIR=$(STAGING_DIR) install
> >> -	for binary in $(LVM2_SBIN); do echo $$binary; cp -a $(STAGING_DIR)/sbin/$$binary $(TARGET_DIR)/sbin; done
> >> +	for binary in $(LVM2_SBIN); do echo $$binary; install -m0755 -p $(STAGING_DIR)/sbin/$$binary $(TARGET_DIR)/sbin; done
>
> Bernhard> close.
> Bernhard> Correct would have been to use $(INSTALL). Also the for
> Bernhard> loop doesn't make too much sense, fwiw.
>
>What does this hunk have to do with $SUBJ ?

No idea, perhaps Nigel would know ;)
It just cought my eye since that's exactly what is just ugly and really,
really should be cleaned up. Folks have to graps that we they are
writing *makefiles* and not shell-scripts in this context.

whatever. I see Nigel just applied it. Nigel, you should perhaps wait at
least a few hours or a day for people to comment, i guess. But then i
don't really care.

[Buildroot] gcc 4.2.2 patch

[Nigel Kukard]

> > Bernhard> On Fri, Mar 28, 2008 at 08:22:02PM +0000, Nigel Kukard wrote:
> > >> Is this patch ok to bump gcc to 4.2.2?
> > >> 
> > >> I grep -r'd the buildroot tree and replaced each occurrence manually and
> > >> removed 4.2.0.
> > >> 
> > >> -N
> >
> > >> Index: package/lvm2/lvm2.mk
> > >> ===================================================================
> > >> --- package/lvm2/lvm2.mk	(revision 21539)
> > >> +++ package/lvm2/lvm2.mk	(working copy)
> > >> @@ -64,7 +64,7 @@
> > >> $(LVM2_TARGET_SBINS): $(LVM2_DIR)/.configured
> > >> $(MAKE) CC=$(TARGET_CC) -C $(LVM2_DIR) DESTDIR=$(STAGING_DIR)
> > >> $(MAKE) CC=$(TARGET_CC) -C $(LVM2_DIR) DESTDIR=$(STAGING_DIR) install
> > >> -	for binary in $(LVM2_SBIN); do echo $$binary; cp -a $(STAGING_DIR)/sbin/$$binary $(TARGET_DIR)/sbin; done
> > >> +	for binary in $(LVM2_SBIN); do echo $$binary; install -m0755 -p $(STAGING_DIR)/sbin/$$binary $(TARGET_DIR)/sbin; done
> >
> > Bernhard> close.
> > Bernhard> Correct would have been to use $(INSTALL). Also the for
> > Bernhard> loop doesn't make too much sense, fwiw.
> >
> >What does this hunk have to do with $SUBJ ?

Nothing at all ... I was being a twit again. Just cleaned up and created
a nice clean svn repo to apply patches against to commit  :)

> I see Nigel just applied it. Nigel, you should perhaps wait at
> least a few hours or a day for people to comment, i guess. But then i
> don't really care.

Applied the GCC version change ... just wanted at least one other person
to see if it looks ok before committing, went over it a few times and it
looked ok, applied it to my svn repo and it got mangled with a lvm2
bugfix I'm busy with. Didn't commit the LVM2 change  :)

-N


-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: application/pgp-signature
Size: 189 bytes
Desc: This is a digitally signed message part
Url : http://busybox.net/lists/buildroot/attachments/20080328/8a06be36/attachment.pgp 

[Buildroot] gcc 4.2.2 patch

[Nigel Kukard]

> >Is this patch ok to bump gcc to 4.2.2?
> >
> >I grep -r'd the buildroot tree and replaced each occurrence manually and
> >removed 4.2.0.
> >
> >-N
> 
> >Index: package/lvm2/lvm2.mk
> >===================================================================
> >--- package/lvm2/lvm2.mk	(revision 21539)
> >+++ package/lvm2/lvm2.mk	(working copy)
> >@@ -64,7 +64,7 @@
> > $(LVM2_TARGET_SBINS): $(LVM2_DIR)/.configured
> > 	$(MAKE) CC=$(TARGET_CC) -C $(LVM2_DIR) DESTDIR=$(STAGING_DIR)
> > 	$(MAKE) CC=$(TARGET_CC) -C $(LVM2_DIR) DESTDIR=$(STAGING_DIR) install
> >-	for binary in $(LVM2_SBIN); do echo $$binary; cp -a $(STAGING_DIR)/sbin/$$binary $(TARGET_DIR)/sbin; done
> >+	for binary in $(LVM2_SBIN); do echo $$binary; install -m0755 -p $(STAGING_DIR)/sbin/$$binary $(TARGET_DIR)/sbin; done

Dammit ... I must stop diff'ing multiple changes! or I must stop
drinking while diff'ing  ;)

I'll commit the gcc bump above.

> close.
> Correct would have been to use $(INSTALL). Also the for loop doesn't
> make too much sense, fwiw.
> 
> Something like this should do everything that is needed and is not
> bloated:
> 	$(INSTALL) -D -m0755 $(patsubst %,$(STAGING_DIR)/sbin/%,$(LVM2_SBIN)) $(TARGET_DIR)/sbin/
> 
> Remember that this is *make* and not your favourite shell..

Thanks, I'll keep that in mind. I've found quite a number of for's in
the tree, if I come across them again I'll fix it.

-N
-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: application/pgp-signature
Size: 189 bytes
Desc: This is a digitally signed message part
Url : http://busybox.net/lists/buildroot/attachments/20080328/0be9b26e/attachment.pgp 

[Buildroot] gcc 4.2.2 patch

[Bernhard Fischer]

On Fri, Mar 28, 2008 at 09:03:42PM +0000, Nigel Kukard wrote:
>> close.
>> Correct would have been to use $(INSTALL). Also the for loop doesn't
>> make too much sense, fwiw.
>> 
>> Something like this should do everything that is needed and is not
>> bloated:
>> 	$(INSTALL) -D -m0755 $(patsubst %,$(STAGING_DIR)/sbin/%,$(LVM2_SBIN)) $(TARGET_DIR)/sbin/
>> 
>> Remember that this is *make* and not your favourite shell..
>
>Thanks, I'll keep that in mind. I've found quite a number of for's in
>the tree, if I come across them again I'll fix it.

Fair enough, but it would be nicer if you would take it upon you to fix
them in one sweep and once and for all. Every such occurance should be
fixed and no new occurance (or touch thereof) should be accepted into
the tree.
Same with all those (="??%?) touch /the/make_target. There is $@ and it
should be used for the sake of cleanlyness. I could go on with another
handful of such cruft but you get the idea..