On 11/30/22 17:50, Kim Phillips wrote:
> @@ -1240,8 +1240,11 @@ static const struct {
> { "retpoline,lfence", SPECTRE_V2_CMD_RETPOLINE_LFENCE, false },
> { "retpoline,generic", SPECTRE_V2_CMD_RETPOLINE_GENERIC, false },
> { "eibrs", SPECTRE_V2_CMD_EIBRS, false },
> + { "autoibrs", SPECTRE_V2_CMD_EIBRS, false },
> { "eibrs,lfence", SPECTRE_V2_CMD_EIBRS_LFENCE, false },
> + { "autoibrs,lfence", SPECTRE_V2_CMD_EIBRS_LFENCE, false },
> { "eibrs,retpoline", SPECTRE_V2_CMD_EIBRS_RETPOLINE, false },
> + { "autoibrs,retpoline", SPECTRE_V2_CMD_EIBRS_RETPOLINE, false },
> { "auto", SPECTRE_V2_CMD_AUTO, false },
> { "ibrs", SPECTRE_V2_CMD_IBRS, false },
I don't think we should expose "autoibrs" to end users like this.
"eibrs" means always-on IBRS. Intel did it first, so gets to name it.
Those are the rules, and it's why we call it "x86_64" and not whatever
Intel's silly name for it was.
Also, expanding the strings:
> + [SPECTRE_V2_EIBRS_RETPOLINE] = "Mitigation: Enhanced / Automatic IBRS + Retpolines",
is fine, but adding new user-visible options that we have to document is
not.
