[armel / 4.6 regression] could not find a spill register

builds successfully in Debian unstable, so either Linaro or armv7/thumb related?

Thank you for the bug report. I've confirmed this with gcc-linaro-4.6-2011.06-0 on ARM:

michaelh@ursa2:~/linaro/bugs$ /tools/toolchains/arch/armv7l/gcc-linaro-4.6-2011.06-0-armv7l-maverick-cbuild130-ursa3-cortexa9r1/bin/g++ -O2 -c ValueRecords.ii
../../../src/share/native/sun/font/layout/ValueRecords.cpp: In member function 'void ValueRecord::adjustPosition(ValueFormat, const char*, GlyphIterator&, const LEFontInstance*) const':
../../../src/share/native/sun/font/layout/ValueRecords.cpp:162:1: error: could not find a spill register
(insn 210 212 214 13 (set (subreg:SI (reg:HI 463) 0)
        (fix:SI (fix:SF (reg:SF 0 r0)))) ../../../src/share/native/sun/font/layout/ValueRecords.cpp:112 668 {*truncsisf2_vfp}
     (expr_list:REG_DEAD (reg:SF 0 r0)
        (nil)))
../../../src/share/native/sun/font/layout/ValueRecords.cpp:162:1: internal compiler error: in failed_reload, at reload1.c:6011

This is a VFP specific fault. One work-around is to compile with -mfloat-abi=soft. The fault does not exist in gcc-linaro-4.5-2011.06-0, gcc-4.5.3, gcc-4.6.1, or GCC trunk.

The fault was exposed by r106753. The fault existed in trunk from at least r172224 until r174044 but was gone by r174684.

I've set it to high priority as it is a ftbfs, occurs at all optimisation levels, and is in a significant package.

It also seems to happen with libkdcraw, so it's not just this one package. https://launchpad.net/ubuntu/+source/libkdcraw/4:4.6.90+repack1-0ubuntu1/+build/2616014/+files/buildlog_ubuntu-oneiric-armel.libkdcraw_4%3A4.6.90%2Brepack1-0ubuntu1_FAILEDTOBUILD.txt.gz

This looks similar to PR48808/48792, which was fixed by rsandiford in upstream...

The workaround did produce a buildable libkdcraw package:

https://launchpad.net/ubuntu/+source/libkdcraw/4:4.6.90+repack1-0ubuntu2

This bug was fixed in the package gcc-4.6 - 4.6.1-4ubuntu1

---------------
gcc-4.6 (4.6.1-4ubuntu1) oneiric; urgency=low

  * Merge with Debian.

gcc-4.6 (4.6.1-4) unstable; urgency=low

  * Update to SVN 20110714 (r176280) from the gcc-4_6-branch.
    - Fix PR tree-optimization/49094, PR target/39633, PR c++/49672,
      PR fortran/49698, PR fortran/49690, PR fortran/49562, PR libfortran/49296,
      PR target/49487, PR tree-optimization/49651, PR ada/48711.

  [ Matthias Klose ]
  * Build Go on alpha for gcc-snapshot builds.
  * For multicore ARM, clear both caches, not just the dcache (proposed
    patch by Andrew Haley).
  * Fix for PR rtl-optimization/{48830,48808,48792}, taken from the trunk.
    LP: #807573.
  * Fix PR tree-optimization/49169, optimisations strip the Thumb/ARM mode bit
    off function pointers (Richard Sandiford). LP: #721531.

  [ Marcin Juszkiewicz ]
  * Define DEB_TARGET_MULTIARCH macro.
  * debian/rules2: Macro and configuration consolidation.

gcc-4.6 (4.6.1-3) unstable; urgency=medium

  * Update to SVN 20110709 (r176108) from the gcc-4_6-branch.
    - Fix PR target/49335, PR tree-optimization/49618, PR c++/49598,
      PR fortran/49479, PR target/49621, PR target/46779, PR target/49660,
      PR c/49644, PR debug/49522, PR debug/49522, PR middle-end/49640,
      PR c++/48157, PR c/49644, PR fortran/48926.
    - Apparently fixes a boost issue. Closes: #632938.
  * Apply proposed patch for PR fortran/49690. Closes: #631204.

  * README.Debian: New section 'Former and/or inactive maintainers'.
 -- Matthias Klose <email address hidden> Thu, 14 Jul 2011 20:18:27 +0200

the work around can be removed

This bug was fixed in the package libkdcraw - 4:4.6.90+repack1-0ubuntu3

---------------
libkdcraw (4:4.6.90+repack1-0ubuntu3) oneiric; urgency=low

  * Drop work around for LP: #807573 now that it's fixed
 -- Scott Kitterman <email address hidden> Fri, 22 Jul 2011 10:04:50 -0400