[OE-core] [PATCH 1/1] opensp: Fix hard path in native nsgmls.
Zhai, Edwin
edwin.zhai at intel.com
Fri Sep 9 14:16:58 UTC 2011
On 2011/9/9 12:49, Saul Wold wrote:
> On 09/08/2011 07:53 PM, edwin.zhai at intel.com wrote:
>> From: Zhai Edwin<edwin.zhai at intel.com>
>>
>> This relocation issues caused following bug and introduce one partial fix bb024d:
>> [YOCTO #1440]
>>
>> But the environment val hacking is not working all the time, as nsgmsl may
>> append wrong path in some case. This fix use search directory "-D" as solution.
>> It's better if got fixed from opensp side when next upgrade.
>>
>> Signed-off-by: Zhai Edwin<edwin.zhai at intel.com>
>> CC: Saul Would<sgw at linux.intel.com>
>> ---
>> meta/recipes-devtools/opensp/opensp_1.5.bb | 4 ++--
>> 1 files changed, 2 insertions(+), 2 deletions(-)
>>
>> diff --git a/meta/recipes-devtools/opensp/opensp_1.5.bb b/meta/recipes-devtools/opensp/opensp_1.5.bb
>> index 0269c87..d692137 100644
>> --- a/meta/recipes-devtools/opensp/opensp_1.5.bb
>> +++ b/meta/recipes-devtools/opensp/opensp_1.5.bb
>> @@ -44,8 +44,8 @@ do_install_append() {
>>
>> do_install_append_virtclass-native() {
>> for util in nsgmls sgmlnorm spam spcat spent sx; do
>> - create_wrapper ${D}/${bindir}/$util \
>> - SGML_CATALOG_FILES=${sysconfdir}/sgml/catalog
>> + create_cmdline_wrapper ${D}/${bindir}/$util \
>> + -D ${sysconfdir}/sgml
>> done
>> }
>>
> Have you tested this with iputils? When I tried this, it did not work
> correctly for me, can you give me more of idea of how you tested this.
Sorry, I forgot to rename the tmp to make sysroot unavailable:( After
renaming, the bug still exist.
Upgrading to 1.5.2 doesn't help. After debug with strace, I found nsgmls
append ".." to other paths when inputting "../index.db" as you said.
It's reasonable for entity such as ping.sgml, as they usually sit in
same dir as index.db. But catalog should not do this.
I have tried "-c" to specify catalog path, and found it works. Could
you pls. have a try on the same contrib tree?
contrib/gzhai/master2
Thanks,
Edwin
>
> Sau!
>
More information about the Openembedded-core
mailing list