On Wed, Jul 9, 2014 at 7:13 PM, Diego Novillo <[email protected]> wrote: > On Wed, Jul 9, 2014 at 10:10 PM, Diego Novillo <[email protected]> wrote: >> On Wed, Jul 9, 2014 at 9:35 PM, Nico Weber <[email protected]> wrote: >>> Did you measure the compile time impact of this? >> >> Yes, in the message I mentioned in my patch description: >> >>>> >>>> http://lists.cs.uiuc.edu/pipermail/llvmdev/2014-June/074290.html > > Pushed send too soon. Sorry. > > The compile time effect is nil. The effect on generated debug info > size is ~1% (less than 1% overall) and approximately 3% for -gmlt.
Just to clarify, this is % of total debug info size (all .debug_* (and .reloc.debug_*) sections, etc) not the % of total object/dwo size, right? > I > tested this over our whole internal code base. > > > Diego. _______________________________________________ cfe-commits mailing list [email protected] http://lists.cs.uiuc.edu/mailman/listinfo/cfe-commits
