Re: [PATCHv8 02/34] lib/vdso: make do_hres and do_coarse as __always_inline

Vincenzo Frascino [off-list ref] writes:

On 11/12/19 1:26 AM, Dmitry Safonov wrote:

quoted

+» » vd·=·&vd[CS_HRES_COARSE];
+out_hres:
+» » return·do_hres(vd,·clock,·ts);
» }·else·if·(msk·&·VDSO_COARSE)·{
» » do_coarse(&vd[CS_HRES_COARSE],·clock,·ts);
» » return·0;
» }·else·if·(msk·&·VDSO_RAW)·{
-» » return·do_hres(&vd[CS_RAW],·clock,·ts);
+» » vd·=·&vd[CS_RAW];
+» » /*·goto·allows·to·avoid·extra·inlining·of·do_hres.·*/
+» » goto·out_hres;

What is the performance impact of "goto out_hres"?

On x86 it's invisible at least in my limited testing.

Thanks,

        tglx