[OE-core] [PATCH 1/1] rpm: fix for N32 MIPS64

[Mark Hatle]

On 11/26/15 4:35 AM, Robert Yang wrote:
> It is 'N32 MIPS64', not 'N32 MIPS32' as command file shows:
> $ file image/usr/bin/getent
> getent: ELF 32-bit MSB executable, MIPS, N32 MIPS64 [snip]
> 
> And "rpm -qp --filecolor" was wrong (it was 1, but should be 4), which
> caused multilib installation error.
> 
> Signed-off-by: Robert Yang <liezhi.yang at windriver.com>
> ---
>  .../rpm/rpm/rpm-rpmfc.c-fix-for-N32-MIPS64.patch   |   34 ++++++++++++++++++++
>  meta/recipes-devtools/rpm/rpm_5.4+cvs.bb           |    1 +
>  meta/recipes-devtools/rpm/rpm_5.4.14.bb            |    1 +
>  3 files changed, 36 insertions(+)
>  create mode 100644 meta/recipes-devtools/rpm/rpm/rpm-rpmfc.c-fix-for-N32-MIPS64.patch
> 
> diff --git a/meta/recipes-devtools/rpm/rpm/rpm-rpmfc.c-fix-for-N32-MIPS64.patch b/meta/recipes-devtools/rpm/rpm/rpm-rpmfc.c-fix-for-N32-MIPS64.patch
> new file mode 100644
> index 0000000..a516574
> --- /dev/null
> +++ b/meta/recipes-devtools/rpm/rpm/rpm-rpmfc.c-fix-for-N32-MIPS64.patch
> @@ -0,0 +1,34 @@
> +From 16dc683aa50be9789d1674734b06a8a955ff22ad Mon Sep 17 00:00:00 2001
> +From: Robert Yang <liezhi.yang at windriver.com>
> +Date: Thu, 26 Nov 2015 01:36:40 -0800
> +Subject: [PATCH] lib/rpmfc.c: fix for N32 MIPS64
> +
> +It is 'N32 MIPS64', not 'N32 MIPS32' as command file shows:
> +$ file image/usr/bin/getent
> +getent: ELF 32-bit MSB executable, MIPS, N32 MIPS64 [snip]
> +
> +And "rpm -qp --filecolor" was wrong (it was 1, but should be 4).
> +
> +Upstream-Status: Pending
> +
> +Signed-off-by: Robert Yang <liezhi.yang at windriver.com>
> +---
> + lib/rpmfc.c |    2 +-
> + 1 file changed, 1 insertion(+), 1 deletion(-)
> +
> +diff --git a/lib/rpmfc.c b/lib/rpmfc.c
> +index fde00c7..49779f8 100644
> +--- a/lib/rpmfc.c
> ++++ b/lib/rpmfc.c
> +@@ -575,7 +575,7 @@ static struct rpmfcTokens_s rpmfcTokens[] = {
> +   { " not stripped",		RPMFC_NOTSTRIPPED },
> +   { " archive",			RPMFC_ARCHIVE },
> + 
> +-  { "MIPS, N32 MIPS32",		RPMFC_ELFMIPSN32|RPMFC_INCLUDE },
> ++  { "MIPS, N32 MIPS64",		RPMFC_ELFMIPSN32|RPMFC_INCLUDE },

I mentioned this to Robert off the list as well.  But I believe we need to look
for both N32 MIPS32 and N32 MIPS64 here.

Older versions of file 'magic' (which is what rpmfc.c is looking at) reported
'N32 MIPS32', while it appears newer versions report 'N32 MIPS64'.  So checking
both is my recommendation.  But this does fix an issue where you can't install
both N32 and O32 multilibs without ELF collisions.

--Mark

> +   { "ELF 32-bit",		RPMFC_ELF32|RPMFC_INCLUDE },
> +   { "ELF 64-bit",		RPMFC_ELF64|RPMFC_INCLUDE },
> + 
> +-- 
> +1.7.9.5
> +
> diff --git a/meta/recipes-devtools/rpm/rpm_5.4+cvs.bb b/meta/recipes-devtools/rpm/rpm_5.4+cvs.bb
> index 951b251..0d14728 100644
> --- a/meta/recipes-devtools/rpm/rpm_5.4+cvs.bb
> +++ b/meta/recipes-devtools/rpm/rpm_5.4+cvs.bb
> @@ -111,6 +111,7 @@ SRC_URI = "cvs://anonymous@rpm5.org/cvs;tag=rpm-5_4;module=rpm \
>  	   file://rpm-rpmpgp-fix.patch \
>  	   file://rpm-disable-Wno-override-init.patch \
>  	   file://rpm-realpath.patch \
> +	   file://rpm-rpmfc.c-fix-for-N32-MIPS64.patch \
>  	  "
>  
>  # Uncomment the following line to enable platform score debugging
> diff --git a/meta/recipes-devtools/rpm/rpm_5.4.14.bb b/meta/recipes-devtools/rpm/rpm_5.4.14.bb
> index 3930410..8e1d3cb 100644
> --- a/meta/recipes-devtools/rpm/rpm_5.4.14.bb
> +++ b/meta/recipes-devtools/rpm/rpm_5.4.14.bb
> @@ -101,6 +101,7 @@ SRC_URI = "http://www.rpm5.org/files/rpm/rpm-5.4/rpm-5.4.14-0.20131024.src.rpm;e
>  	   file://rpm-opendb-before-verifyscript-to-avoid-null-point.patch \
>  	   file://configure.ac-check-for-both-gpg2-and-gpg.patch \
>  	   file://0001-define-EM_AARCH64.patch \
> +	   file://rpm-rpmfc.c-fix-for-N32-MIPS64.patch \
>  	  "
>  
>  # Uncomment the following line to enable platform score debugging
>