On Thu, Sep 20, 2018 at 01:16:09PM +0100, Emil Velikov wrote: > On 14 September 2018 at 00:57, Lucas De Marchi <lucas.demarchi@xxxxxxxxx> wrote: > > Rely on -fvisibility=hidden to hide the symbols. Previous version of > > this series applying only to drm_intel.so is > > > > Reviewed-by: Eric Engestrom <eric.engestrom@xxxxxxxxx> > > > > but it's not included here since I changed the approach for the build > > system change. > > > > drm_private can also be removed from other symbols but it proved to be > > a lot of manual work to re-align all the fields, so I decided to leave > > it to be done on top as a cleanup. > > > Did you read through the reasoning behind > 0f8da82500ec542e269092c0718479e25eaff5f6? In v1 (https://lists.freedesktop.org/archives/dri-devel/2018-September/189639.html) I had this: From git log archeology and mention in another thread we used to pass -fvisibility=hidden, but reverted to the contrary due to bug in obscure toolchain some years ago (see 0f8da82500ec542e269092c0718479e25eaff5f6). I think it's time to revisit that decision: we have plenty of other projects doing that nowadays without problem. It looks like for some reason this was not part of the coverletter in v2, sorry. > > Alternatively you do not care too much about compat in said cases, > which is fine. > My only request is - keep it only for the Intel bits. this is not only for Intel bits, it's improving all the libdrm as a whole. A decision/commit made 3 year can be revisited, we don't have to carry it forever. Do we have a problem with it *today*? If we do, then we may have another decision like working it around for the sad said compiler rather than for everybody. Default visibility hidden is a sane default adopted by multiple open source libraries as it avoids unwanted results that can't be taken back (like exporting a symbol that shouldn't have been) and reduces maintenance (just look at the custom scripts we have to keep checking the exported symbols). Lucas De Marchi > > Thanks > Emil _______________________________________________ dri-devel mailing list dri-devel@xxxxxxxxxxxxxxxxxxxxx https://lists.freedesktop.org/mailman/listinfo/dri-devel
