[Buildroot] [PATCH v2,1/1] package/dmalloc: needs -fPIC
Peter Korsgaard
peter at korsgaard.com
Fri May 7 07:32:25 UTC 2021
>>>>> "Arnout" == Arnout Vandecappelle <arnout at mind.be> writes:
> On 17/04/2021 12:49, Fabrice Fontaine wrote:
> [snip]
>> diff --git a/package/dmalloc/0001-add-fpic.patch b/package/dmalloc/0001-add-fpic.patch
>> deleted file mode 100644
>> index 230a9e15f1..0000000000
>> --- a/package/dmalloc/0001-add-fpic.patch
>> +++ /dev/null
>> @@ -1,58 +0,0 @@
>> -Without -fPIC, dmalloc does not build with:
>> -
>> -/home/test/mips-4.4/bin/mips-linux-gnu-ld
>> --sysroot=/home/test/outputs/test-35/staging -shared --whole-archive
>> -soname libdmallocxx.so -o libdmallocxx.so.t libdmallocxx.a
>> -/home/test/mips-4.4/bin/mips-linux-gnu-ld:
>> libdmalloc.a(arg_check.o): relocation R_MIPS_HI16 against
>> `_dmalloc_flags' can not be used when making a shared object;
>> recompile with -fPIC
>> -libdmalloc.a(arg_check.o): could not read symbols: Bad value
>> -
>> -This patch, taken from
>> -http://sources.gentoo.org/cgi-bin/viewvc.cgi/gentoo-x86/dev-libs/dmalloc/files/dmalloc-5.2.4-fpic.patch?hideattic=0&view=markup,
>> -fixes the problem by passing the -fPIC flag. It isn't passed through
>> -the ./configure environment in order to not clutter the configuration
>> -cache with incorrect values.
> I was going to apply, and then I read the sentence above and I wasn't so sure.
> I don't know though what Thomas means with "in order not to clutter the
> configuration cache with incorrect values"... Thomas, it's just 11 years ago,
> I'm sure you still remember the reasoning behind it?
> Oh what the heck, this is useless anyway. I've applied after all. Thomas, in
> the unlikely case that you *do* remember and that this fix is wrong, please
> revert 19ec872f169a851b48ba04d22432b7c0939847d4. Otherwise, the autobuilders
> will let us know.
I *think* this was related to back when we tried to use a shared
config.cache between packages, E.G. commit 5432f26f0e78fc74a5.
Committed to 2021.02.x, thanks.
--
Bye, Peter Korsgaard
More information about the buildroot
mailing list