https://patchwork.ozlabs.org/patch/503550/
binutils: fix buildng of Linux kernel for ARCv2 ISA
Submitted by
Alexey Brodkin on Aug. 4, 2015, 11:42 a.m.
Details
| Message ID | 1438688562-26476-1-git-send-email-abrodkin@synopsys.com |
|---|---|
| State | Accepted |
| Headers | show |
Commit Message
Comments
Dear Alexey Brodkin, On Tue, 4 Aug 2015 14:42:42 +0300, Alexey Brodkin wrote: > With fix of missing .tdata/.tbss sections we unintentionally introduced > a regression for ARCv2 ISA (read ARC HS38) kernel building. > > That's what we got on attempt to build kernel: > ----------------------------------->8-------------------------------------- > LD drivers/video/fbdev/built-in.o > arc-linux-ld: ERROR: Attempting to link drivers/video/fbdev/omap2/built-in.o with a binary drivers/video/fbdev/built-in.o of different architecture > arc-linux-ld: failed to merge target specific data of file drivers/video/fbdev/omap2/built-in.o > scripts/Makefile.build:337: recipe for target 'drivers/video/fbdev/built-in.o' failed > make[3]: *** [drivers/video/fbdev/built-in.o] Error 1 > scripts/Makefile.build:403: recipe for target 'drivers/video/fbdev' failed > make[2]: *** [drivers/video/fbdev] Error 2 > scripts/Makefile.build:403: recipe for target 'drivers/video' failed > make[1]: *** [drivers/video] Error 2 > Makefile:944: recipe for target 'drivers' failed > make: *** [drivers] Error 2 > ----------------------------------->8-------------------------------------- I've fixed the typo in the commit title, and applied. Thanks! Thomas
Patch hide | download patch | download mbox
diff --git a/package/binutils/arc-2015.06/0004-Check-to-see-if-the-input-BFD-actually-contains-any-.patch b/package/binutils/arc-2015.06/0004-Check-to-see-if-the-input-BFD-actually-contains-any-.patch new file mode 100644 index 0000000..fe4808f --- /dev/null +++ b/package/binutils/arc-2015.06/0004-Check-to-see-if-the-input-BFD-actually-contains-any-.patch @@ -0,0 +1,132 @@ +From a65b844aed9153789356e098984452df2f5d9058 Mon Sep 17 00:00:00 2001 +From: Claudiu Zissulescu <claziss@synopsys.com> +Date: Tue, 4 Aug 2015 12:53:11 +0200 +Subject: [PATCH] Check to see if the input BFD actually contains any sections. + +--- + bfd/elf32-arc.c | 70 +++++++++++++++++++++++++++++++++++++------------------ + 1 file changed, 47 insertions(+), 23 deletions(-) + +diff --git a/bfd/elf32-arc.c b/bfd/elf32-arc.c +index 38f72b4..76bac6c 100644 +--- a/bfd/elf32-arc.c ++++ b/bfd/elf32-arc.c +@@ -881,44 +881,68 @@ arc_elf_merge_private_bfd_data (bfd *ibfd, bfd *obfd) + { + unsigned short mach_ibfd; + static unsigned short mach_obfd = EM_NONE; +- flagword old_flags; +- flagword new_flags; ++ flagword out_flags; ++ flagword in_flags; ++ asection *sec; ++ ++ /* Check if we have the same endianess. */ ++ if (! _bfd_generic_verify_endian_match (ibfd, obfd)) ++ { ++ _bfd_error_handler ( ++ _("ERROR: Endian Match failed . Attempting to link %B with binary %s ++of opposite endian-ness"), ++ ibfd, bfd_get_filename (obfd)); ++ return FALSE; ++ } + + /* Collect ELF flags. */ +- new_flags = elf_elfheader (ibfd)->e_flags & EF_ARC_MACH_MSK; +- old_flags = elf_elfheader (obfd)->e_flags & EF_ARC_MACH_MSK; ++ in_flags = elf_elfheader (ibfd)->e_flags & EF_ARC_MACH_MSK; ++ out_flags = elf_elfheader (obfd)->e_flags & EF_ARC_MACH_MSK; + + #if DEBUG +- (*_bfd_error_handler) ("old_flags = 0x%.8lx, new_flags = 0x%.8lx, init = %s, filename = %s", +- old_flags, new_flags, elf_flags_init (obfd) ? "yes" : "no", ++ (*_bfd_error_handler) ("out_flags = 0x%.8lx, in_flags = 0x%.8lx, init = %s, filename = %s", ++ out_flags, in_flags, elf_flags_init (obfd) ? "yes" : "no", + bfd_get_filename (ibfd)); + #endif + + if (!elf_flags_init (obfd)) /* First call, no flags set. */ + { + elf_flags_init (obfd) = TRUE; +- old_flags = new_flags; ++ out_flags = in_flags; + } + + if (bfd_get_flavour (ibfd) != bfd_target_elf_flavour + || bfd_get_flavour (obfd) != bfd_target_elf_flavour) + return TRUE; + +- if (bfd_count_sections (ibfd) == 0) +- return TRUE ; /* For the case of empty archive files */ ++ /* Check to see if the input BFD actually contains any sections. If ++ not, its flags may not have been initialised either, but it ++ cannot actually cause any incompatiblity. Do not short-circuit ++ dynamic objects; their section list may be emptied by ++ elf_link_add_object_symbols. */ ++ if (!(ibfd->flags & DYNAMIC)) ++ { ++ bfd_boolean null_input_bfd = TRUE; ++ bfd_boolean only_data_sections = TRUE; + +- mach_ibfd = elf_elfheader (ibfd)->e_machine; ++ for (sec = ibfd->sections; sec != NULL; sec = sec->next) ++ { ++ if ((bfd_get_section_flags (ibfd, sec) ++ & (SEC_LOAD | SEC_CODE | SEC_HAS_CONTENTS)) ++ == (SEC_LOAD | SEC_CODE | SEC_HAS_CONTENTS)) ++ only_data_sections = FALSE; + +- /* Check if we have the same endianess. */ +- if (! _bfd_generic_verify_endian_match (ibfd, obfd)) +- { +- _bfd_error_handler (_(" +-ERROR: Endian Match failed . Attempting to link %B with binary %s +-of opposite endian-ness"), +- ibfd, bfd_get_filename (obfd)); +- return FALSE; ++ null_input_bfd = FALSE; ++ break; ++ } ++ ++ if (null_input_bfd || only_data_sections) ++ return TRUE; + } + ++ ++ /* Complain about various flag/architecture mismatches. */ ++ mach_ibfd = elf_elfheader (ibfd)->e_machine; + if (mach_obfd == EM_NONE) + { + mach_obfd = mach_ibfd; +@@ -932,23 +956,23 @@ with a binary %s of different architecture"), + ibfd, bfd_get_filename (obfd)); + return FALSE; + } +- else if (new_flags != old_flags) ++ else if (in_flags != out_flags) + { + /* Warn if different flags. */ + (*_bfd_error_handler) + (_("%s: uses different e_flags (0x%lx) fields than previous modules (0x%lx)"), +- bfd_get_filename (ibfd), (long)new_flags, (long)old_flags); +- if (new_flags && old_flags) ++ bfd_get_filename (ibfd), (long)in_flags, (long)out_flags); ++ if (in_flags && out_flags) + return FALSE; + /* MWDT doesnt set the eflags hence make sure we choose the + eflags set by gcc. */ +- new_flags = new_flags > old_flags ? new_flags : old_flags; ++ in_flags = in_flags > out_flags ? in_flags : out_flags; + } + + } + + /* Update the flags. */ +- elf_elfheader (obfd)->e_flags = new_flags; ++ elf_elfheader (obfd)->e_flags = in_flags; + + if (bfd_get_mach (obfd) < bfd_get_mach (ibfd)) + { +-- +2.4.3 +
With fix of missing .tdata/.tbss sections we unintentionally introduced a regression for ARCv2 ISA (read ARC HS38) kernel building. That's what we got on attempt to build kernel: ----------------------------------->8-------------------------------------- LD drivers/video/fbdev/built-in.o arc-linux-ld: ERROR: Attempting to link drivers/video/fbdev/omap2/built-in.o with a binary drivers/video/fbdev/built-in.o of different architecture arc-linux-ld: failed to merge target specific data of file drivers/video/fbdev/omap2/built-in.o scripts/Makefile.build:337: recipe for target 'drivers/video/fbdev/built-in.o' failed make[3]: *** [drivers/video/fbdev/built-in.o] Error 1 scripts/Makefile.build:403: recipe for target 'drivers/video/fbdev' failed make[2]: *** [drivers/video/fbdev] Error 2 scripts/Makefile.build:403: recipe for target 'drivers/video' failed make[1]: *** [drivers/video] Error 2 Makefile:944: recipe for target 'drivers' failed make: *** [drivers] Error 2 ----------------------------------->8-------------------------------------- The reason was in empty .tdata and .tbss sections in empty archives. And later empty archives were linked in built-in.o with default architecture (in our case ARCv1 ISA, read for ARC 700) and then expected failure happened when objets for different architectures were attempted to link together. Now we have a fix for that issue, see https://github.com/foss-for-synopsys-dwc-arc-processors/binutils-gdb/commit/a65b844aed9153789356e098984452df2f5d9058 This fix is in arc-2.23-dev branch and will be a part of the next release of ARC tools, so then this patch must be removed from buildroot. Signed-off-by: Alexey Brodkin <abrodkin@synopsys.com> Cc: Thomas Petazzoni <thomas.petazzoni@free-electrons.com> --- ...e-if-the-input-BFD-actually-contains-any-.patch | 132 +++++++++++++++++++++ 1 file changed, 132 insertions(+) create mode 100644 package/binutils/arc-2015.06/0004-Check-to-see-if-the-input-BFD-actually-contains-any-.patch