On Fri, Jun 15, 2012 at 5:55 AM, Richard Purdie
<richard.pur...@linuxfoundation.org> wrote:
> On Thu, 2012-06-14 at 23:12 -0700, Khem Raj wrote:
>> futimes is not available on uclibc so use utimes
>> qsort is also not as expected by sat-solver therefore
>> for uclibc we resort to using internal version of
>> qsort
>>
>> Signed-off-by: Khem Raj <raj.k...@gmail.com>
>> ---
>> .../sat-solver/sat-solver/futimes.patch | 32
>> ++++++++++++++++++++
>> meta/recipes-extended/sat-solver/sat-solver_git.bb | 3 ++
>> 2 files changed, 35 insertions(+), 0 deletions(-)
>> create mode 100644 meta/recipes-extended/sat-solver/sat-solver/futimes.patch
>>
>> diff --git a/meta/recipes-extended/sat-solver/sat-solver/futimes.patch
>> b/meta/recipes-extended/sat-solver/sat-solver/futimes.patch
>> new file mode 100644
>> index 0000000..b24d852
>> --- /dev/null
>> +++ b/meta/recipes-extended/sat-solver/sat-solver/futimes.patch
>> @@ -0,0 +1,32 @@
>> +This patch uses utimes instead of futimes for uclibc
>> +since futimes is not available
>> +
>> +Upstream-Status: Pending
>> +Signed-off-by: Khem Raj <raj.k...@gmail.com>
>> +
>> +Index: git/examples/solv.c
>> +===================================================================
>> +--- git.orig/examples/solv.c 2012-06-01 12:06:22.041552848 -0700
>> ++++ git/examples/solv.c 2012-06-01 12:49:17.417677449 -0700
>> +@@ -1027,7 +1027,8 @@
>> + int flags;
>> +
>> + cinfo = repo->appdata;
>> +- if (!(fp = fopen(calccachepath(repo, repoext), "r")))
>> ++ const char* fname = calccachepath(repo, repoext);
>> ++ if (!(fp = fopen(fname, "r")))
>> + return 0;
>> + if (fseek(fp, -sizeof(mycookie), SEEK_END) || fread(mycookie,
>> sizeof(mycookie), 1, fp) != 1)
>> + {
>> +@@ -1068,7 +1069,11 @@
>> + memcpy(cinfo->extcookie, myextcookie, sizeof(myextcookie));
>> + }
>> + if (mark)
>> ++#ifdef __UCLIBC__
>> ++ utimes(fname, 0); /* try to set modification time */
>> ++#else
>> + futimes(fileno(fp), 0); /* try to set modification time */
>> ++#endif
>> + fclose(fp);
>> + return 1;
>> + }
>> diff --git a/meta/recipes-extended/sat-solver/sat-solver_git.bb
>> b/meta/recipes-extended/sat-solver/sat-solver_git.bb
>> index b0382f6..b0daa6a 100644
>> --- a/meta/recipes-extended/sat-solver/sat-solver_git.bb
>> +++ b/meta/recipes-extended/sat-solver/sat-solver_git.bb
>> @@ -20,6 +20,7 @@ SRC_URI =
>> "git://github.com/openSUSE/sat-solver.git;protocol=git \
>> file://sat-solver_core.patch \
>> file://fix_gcc-4.6.0_compile_issue.patch \
>> file://0001-sat_xfopen.c-Forward-port-to-zlib-1.2.6-gzFile.patch
>> \
>> + file://futimes.patch \
>> "
>>
>> S = "${WORKDIR}/git"
>> @@ -28,6 +29,8 @@ EXTRA_OECMAKE += "-DRPM5=RPM5 -DOE_CORE=OE_CORE"
>>
>> EXTRA_OECMAKE += " -DLIB=${@os.path.basename('${libdir}')}"
>>
>> +TUNE_CCARGS_libc-uclibc_append = " -DUSE_OWN_QSORT=1 "
>> +
>
> I suspect you mean TUNE_CCARGS_append_libc-uclibc here?
indeed good catch. This was incorrect and it worked because it overwrote
TUNE_CCARGS and I did not see it because the options I wanted
was still present and it was building.
I have updated the pull tree with this changed.
>
> Cheers,
>
> Richard
>
>
> _______________________________________________
> Openembedded-core mailing list
> Openembedded-core@lists.openembedded.org
> http://lists.linuxtogo.org/cgi-bin/mailman/listinfo/openembedded-core
_______________________________________________
Openembedded-core mailing list
Openembedded-core@lists.openembedded.org
http://lists.linuxtogo.org/cgi-bin/mailman/listinfo/openembedded-core
