From: cotulla@yandex.ua Subject: Re: [DISCUSSION] Hexagon code inside kernel Date: Wed, 20 Feb 2013 05:07:46 +0400 Message-ID: <721691361322466@web24g.yandex.ru> References: <1093431360938494@web22d.yandex.ru> <1433341360978513@web2h.yandex.ru> <1163031361018389@web26d.yandex.ru> <20130219043605.GB12127@codeaurora.org> Mime-Version: 1.0 Content-Transfer-Encoding: QUOTED-PRINTABLE Return-path: DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=yandex.ru; s=mail; t=1361322467; bh=fyT9XQVJ6zjpZp6yRQ2wv6f04J8n/gjld74UeZiEnCw=; h=From:To:In-Reply-To:References:Subject:Date; b=XEURyNdt6hhOk3jc4WinRMBqnRUwoTiiFjPtsCE/v/VE0w7HG7zXsSVGognHkgaFi wT9N5P4a2R1foi/jQ4+aSsxu2xx4+qN8jL65DwP1XjQvRtG5NAia7Uu84TxLaeN/SB B0FnJPz8FZwA+Aitj9WAawHH2A72xBbqbS0Ikv64= In-Reply-To: Sender: linux-hexagon-owner@vger.kernel.org List-ID: Content-Type: text/plain; charset="windows-1252" To: linux-hexagon@vger.kernel.org > =9AI might be confused, but I believe that this is actually very comm= on; > =9Athe difference is that the so-called 'hardware MMU''s are actually > =9Ajust nanocode burned into a ROM that's on the cpu chip die. =9AIts > =9Ainvisible to the user. FWIW some deprecated instructions are also > =9Aemulated this way: they fault to a handler. You can hide a lot of > =9Astuff this way. The powerpc does this, they learned it from the > =9Amainframe 370/390 architecture, which abuses this (did you really > =9Athink the 'startio' instruction, which behaves like a DMA device > =9Adriver, was actually a single cycle hardware instruction?) > > =9AThe hexagon doesn't have a (self-loading) ROM, so the mini-VM has = to > =9Abe loaded at boot. > I mean that Hexagon has a simular thing like MIPS.=20 Unlike ARM hardware does not fetch page table entries from RAM, it's us= ing standalone TLB entries. In ARM they are also present but usually ever low level code doesn't wo= rk with them, low level code works with page tables in the RAM. -Cotulla