linux-kernel - Re: leave_mm() and lazy TLB mode

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite: Windows password security audit tool. GUI, reports in PDF.

[<prev] [next>] [<thread-prev] [day] [month] [year] [list]

Date:	Sun, 29 Oct 2006 08:18:44 +0800
From:	"Dong Feng" <middle.fengdong@...il.com>
To:	linux-kernel@...r.kernel.org
Subject: Re: leave_mm() and lazy TLB mode

I read source code more and now I understand that the effect of
leave_mm() is not only

1. bypassing the invocation to local_flush_tlb(), but also
2. prevent the curent CPU from receiving any more INVALIDATE_TLB_VECTOR IPI.

Some material makes me understand the lazy TLB mode this way, that is,
a CPU's TLB will not be flushed since it enters lazy TLB mode till it
return to normal TLB mode.

But according to the source code, it seems that a CPU still has to
undergo at most one time TLB flush, that is, the flush caused by
loading CR3 with swapper_pg_dir.

Could any one confirm whether my original understanding and my
corrective understanding is correct? Thank you very much.

2006/10/29, Dong Feng <middle.fengdong@...il.com>:
> There are several places where some comments say leave_mm() disables
> TLB flush on CPUs in lazy TLB mode.
>
> My question may be silly, ...
>
> leave_mm() change the value of CR3 register. In my understanding, that
> causes TLB flush. Does this contradict with the definition of lazy TLB
> mode?
>
-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@...r.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/