From mboxrd@z Thu Jan  1 00:00:00 1970
Received: from eigenstate.org ([206.124.132.107]) by ewsd; Sat Nov 16 23:50:09 EST 2019
Received: from eigenstate.org (localhost [127.0.0.1])
	by eigenstate.org (OpenSMTPD) with ESMTP id 2c7d4e4b
	for <9front@9front.org>;
	Sat, 16 Nov 2019 20:50:08 -0800 (PST)
DKIM-Signature: v=1; a=rsa-sha1; c=relaxed; d=eigenstate.org; h=
	message-id:to:subject:date:from:mime-version:content-type
	:content-transfer-encoding; s=mail; bh=amA7oo1HayqZMyfFPSXpU18Ot
	x4=; b=MHnl1mguYpoYS5SuHRCm9MnPy4hm7CY8kj1iEcXhNq+X60wwMCsk22Hie
	v8ZmaLGlRcwx7zDiMHclssezYH2v2rfn9LmMs7Xg/3Aij4kqw7/EozhO/lDVUnwI
	z7tX/ufvDfU3spq77EUJRadu6PKy+rU/LB5cWqAD2dC+KMyvhQ=
DomainKey-Signature: a=rsa-sha1; c=nofws; d=eigenstate.org; h=message-id
	:to:subject:date:from:mime-version:content-type
	:content-transfer-encoding; q=dns; s=mail; b=bKmeeC7pT9cf+WPEqlz
	n3Farc8G9hZKKnXmqEfRuoqevO8Ww15m+8P9jwqYIEo8mw9T9ui/qro73pzlnjo7
	Orp4/zkQhLlJEB6kJdf+KJbpwNLsqaa2NCq0FqxlCXL51DQIphwtOvEAoSdUyMWV
	IcEOUcjUrok7A18r6B3i1/cY=
Received: from abbatoir.hsd1.ca.comcast.net (c-76-21-119-139.hsd1.ca.comcast.net [76.21.119.139])
	by eigenstate.org (OpenSMTPD) with ESMTPSA id 91eaf8c6 (TLSv1.2:ECDHE-RSA-AES256-SHA:256:NO)
	for <9front@9front.org>;
	Sat, 16 Nov 2019 20:50:08 -0800 (PST)
Message-ID: <E196E7B9A63F3382559A4EBA852969BC@eigenstate.org>
To: 9front@9front.org
Subject: ahci led: reset spins.
Date: Sat, 16 Nov 2019 20:50:07 -0800
From: ori@eigenstate.org
MIME-Version: 1.0
Content-Type: text/plain; charset="UTF-8"
Content-Transfer-Encoding: 8bit
List-ID: <9front.9front.org>
List-Help: <http://lists.9front.org>
X-Glyph: ➈
X-Bullshit: firewall blockchain software 

On one of my machines, there was aa fairly constant load from the
ialed kproc, because it was stuck in a loop resetting the ahci
controller. The controller never came back.

I'm guessing the controller was confused because there are no
drives attached. I added a timeout as a workaround.

The AHCI spec says:

	If the HBA has not cleared GHC.HR to ‘0’ within 1 second of
	software setting GHC.HR to ‘1’, the HBA is in
	a hung or locked state.

And I've tested on a couple of machines, both with and without
drives attached, and it seems to work --  but given that people
seem to think accessing disks is important, more testing should
probably be done before committing:

diff -r c4896008f196 sys/src/9/pc/sdiahci.c
--- a/sys/src/9/pc/sdiahci.c	Fri Nov 15 13:26:25 2019 -0800
+++ b/sys/src/9/pc/sdiahci.c	Sat Nov 16 20:23:26 2019 -0800
@@ -1215,13 +1215,20 @@
 ahciencreset(Ctlr *c)
 {
 	Ahba *h;
+	int ndelay;
 
 	if(c->enctype == Eesb)
 		return 0;
 	h = c->hba;
 	h->emctl |= Emrst;
-	while(h->emctl & Emrst)
+	ndelay = 0;
+	while(h->emctl & Emrst){
+		if(ndelay++ > 1000){
+			print("hung ctlr: %s\n", Tname(c));
+			return -1;
+		}
 		delay(1);
+	}
 	return 0;
 }
 
@@ -1403,7 +1410,8 @@
 	memset(map, 0, sizeof map);
 	for(i = 0; i < niactlr; i++)
 		if(iactlr[i].enctype != 0){
-			ahciencreset(iactlr + i);
+			if(ahciencreset(iactlr + i) == -1)
+				continue;
 			map[i] = 1;
 			j++;
 		}