On Fri, Nov 24, 2017 at 3:39 AM, David Laight <david.lai...@aculab.com> wrote: > From: Andy Lutomirski >> Sent: 24 November 2017 04:33 > ... >> + /* >> + * x86 lacks a near absolute jump, and we can't jump to the real >> + * entry text with a relative jump, so we fake it using retq. >> + */ >> + pushq $entry_SYSCALL_64_after_hwframe >> + retq > > Don't some of the cpus have hardware logic that follows call and return? > The above will break that logic and slow things down (unless it is an > expected sequence).
Indeed, and it makes a *huge* difference. > > Is there a scratch register that can be used for an indirect jump? No, but I can fudge it. > > David >
