Commits to SVN

Randy McMurchy randy at linuxfromscratch.org
Sat Oct 4 08:03:24 PDT 2008


DJ Lucas wrote:
> DJ Lucas wrote:
>> Still there in latest snapshot:
>>
>> [snip known errors]
>>   
> And after applying the copy that Greg suggested:
> 
> [snip the math error]
> 
> The copy command should be: 
> cp -v ../glibc-2.8-20080929/iconvdata/gconv-modules iconvdata
> and be inserted between make and make check.

So, does that mean you did not need the patch Greg uses over at
DIY before using the copy command (see my message to this list
about the subject, and Greg's reply)?

I was just about to commit the Glibc update, but before I do
I want to run the following questions by DJ.

1. I used the 0922 snapshot, and I needed the patch Greg uses
and the copy command to resolve the iconv errors. I'll change
the version to use the 0929 snapshot as you've tested it. I
still need to know if we need the iconv patch from DIY though.

2. You say in another post to ditch the sed for vi_VN.TCVN.
I'll do that along with adding back the stuff to only do
the partial list of locales. Sound okay?

3. In your comment below, you indicate that we can ditch the
math error using this patch:

2008-07-15  Joseph Myers  <joseph at codesourcery.com>

	* sysdeps/i386/fpu/libm-test-ulps: Add inline long double ulps for
	expm1.

Index: sysdeps/i386/fpu/libm-test-ulps
===================================================================
RCS file: /cvs/glibc/libc/sysdeps/i386/fpu/libm-test-ulps,v
retrieving revision 1.44
diff -u -r1.44 libm-test-ulps
--- sysdeps/i386/fpu/libm-test-ulps	15 Jan 2006 17:59:37 -0000	1.44
+++ sysdeps/i386/fpu/libm-test-ulps	15 Jul 2008 14:28:47 -0000
@@ -453,6 +453,10 @@
  ildouble: 8
  ldouble: 8

+# expm1
+Test "expm1 (1) == M_El - 1.0":
+ildouble: 1
+
  # gamma
  Test "gamma (-0.5) == log(2*sqrt(pi))":
  double: 1
@@ -1134,6 +1138,9 @@
  ildouble: 8
  ldouble: 8

+Function: "expm1":
+ildouble: 1
+
  Function: "gamma":
  double: 1
  idouble: 1


Is that correct? It doesn't spawn failures on other tests?
Bottom line is should we include it? Remember we target for
x86 arches.

TIA for your replies, DJ.


> The annexc.out failure has been there forever.  The test-ildoubl.out is 
> new but expected on x86, but not on x86_64.  See the second link from 
> Greg's post on the DIY List:
> 
> http://www.diy-linux.org/pipermail/diy-linux-dev/2008-September/001280.html
> 
> The second patch located at 
> http://sourceware.org/ml/libc-alpha/2008-07/msg00024.html does correct 
> the failure of the idouble test.

-- 
Randy



More information about the lfs-dev mailing list