|
 Home
|
|
|
PR Database Search
|
From www@netbsd.org Sat May 13 16:36:13 2023
Return-Path: <www@netbsd.org>
Received: from mail.netbsd.org (mail.netbsd.org [199.233.217.200])
(using TLSv1.3 with cipher TLS_AES_256_GCM_SHA384 (256/256 bits))
(Client CN "mail.NetBSD.org", Issuer "mail.NetBSD.org CA" (not verified))
by mollari.NetBSD.org (Postfix) with ESMTPS id 121ED1A923C
for <gnats-bugs@gnats.NetBSD.org>; Sat, 13 May 2023 16:36:13 +0000 (UTC)
Message-Id: <20230513163611.B81DE1A923D@mollari.NetBSD.org>
Date: Sat, 13 May 2023 16:36:11 +0000 (UTC)
From: hpaluch@seznam.cz
Reply-To: hpaluch@seznam.cz
To: gnats-bugs@NetBSD.org
Subject: Fix for kernel freeze with "BSY never cleared, status 0x80" with 1TB Seagate SATA HDD
X-Send-Pr-Version: www-1.0
>Number: 57407
>Category: kern
>Synopsis: Fix for kernel freeze with "BSY never cleared, status 0x80" with 1TB Seagate SATA HDD
>Confidential: no
>Severity: serious
>Priority: medium
>Responsible: kern-bug-people
>State: open
>Class: sw-bug
>Submitter-Id: net
>Arrival-Date: Sat May 13 16:40:00 +0000 2023
>Last-Modified: Sun May 14 17:15:01 +0000 2023
>Originator: Henryk Paluch
>Release: 9.3-RELEASE
>Organization:
N/A
>Environment:
From custom kernel (stock kernel freezes on boot):
NetBSD localhost 9.3 NetBSD 9.3 (GENERIC_HP) #0: Sat May 13 08:32:51 UTC 2023 root@netbsd93.example.com:/usr/src/sys/arch/amd64/compile/GENERIC_HP amd64
>Description:
When I boot standard NetBSD 9.3 CD - tested NetBSD-9.3-amd64.iso -
with new 1TB Seagate drive: ST1000DM010-2EP102 the kernel freeze on SATA link reset with message "BSY never cleared, status 0x80".
Here are relevant messages regarding my SATA controller and that HDD:
viaide2 at pci0 dev 5 function 1: NVIDIA MCP55 Serial ATA Controller (rev. 0xa2)
viaide2: bus-master DMA support present
viaide2: primary channel wired to native-PCI mode
viaide2: using ioapic0 pin 21 for native-PCI interrupt
atabus3 at viaide2 channel 0
viaide2: secondary channel wired to native-PCI mode
atabus4 at viaide2 channel 1
viaide2 port 1: device present, speed: 3.0Gb/s
viaide2: BSY never cleared, status 0x80
I have found that other HDD (old WD 500GB SATA disk) works without problem with same NetBSD. Also OpenBSD and Linux boots on same machine without issue (installed in GPT+BIOS with few small hacks).
There one report which can be related - on Internet, for example:
- https://www.unitedbsd.com/d/605-netbsd-92-amd64-does-not-recognize-hard-disk/5
However it is not yet confirmed.
More HW details:
- MB MSI K9N Platinum
- nVidia nForce 570 chipset
- CPU AMD X2 (Dual-Core Athlon)
- 8GB of RAM
- here are details on HDD with fixed kernel:
wd0 at atabus4 drive 0
wd0: <ST1000DM010-2EP102>
wd0: drive supports 16-sector PIO transfers, LBA48 addressing
wd0: 931 GB, 1938021 cyl, 16 head, 63 sec, 512 bytes/sect x 1953525168 sectors (0 bytes/physsect; first aligned sector: 8)
wd0: GPT GUID: 044fdb79-3b9c-4fff-86a9-66becdf9aebd
...
>How-To-Repeat:
Boot NetBSD 9.3 with old SATA controller (tested NVIDIA MCP55 Serial ATA) and new Seagate 1TB SATA disk (tested ST1000DM010-2EP102).
The kernel will freeze for around 5minutes with:
BSY never cleared, status 0x80
After around 5 minutes it will continue boot, however such disk is completely ignored, so NetBSD can't be installed on it.
>Fix:
I found that this trivial patch, that forces SATA1 1.5Gb/s mode fixes the problem and I was able to installed NetBSD on that system:
--- /root/sata_subr.c 2023-05-13 07:38:30.264874166 +0000
+++ /usr/src/sys/dev/ata/sata_subr.c 2023-05-13 07:40:24.075628047 +0000
@@ -93,8 +93,10 @@
* the device. It doesn't hurt for other devices.
*/
bus_space_write_4(sata_t, scontrol_r, 0, 0);
- scontrol = SControl_IPM_NONE | SControl_SPD_ANY | SControl_DET_INIT;
+ scontrol = SControl_IPM_NONE | SControl_SPD_G1 | SControl_DET_INIT;
bus_space_write_4(sata_t, scontrol_r, 0, scontrol);
+ aprint_normal("%s port %d: enforcing SATA1 speed at 1.5Gb/s\n",
+ device_xname(chp->ch_atac->atac_dev), chp->ch_channel);
ata_delay(chp, 50, "sataup", flags);
scontrol &= ~SControl_DET_INIT;
>Audit-Trail:
From: matthew green <mrg@eterna.com.au>
To: gnats-bugs@netbsd.org
Cc: kern-bug-people@netbsd.org, gnats-admin@netbsd.org,
netbsd-bugs@netbsd.org
Subject: re: kern/57407: Fix for kernel freeze with "BSY never cleared, status 0x80" with 1TB Seagate SATA HDD
Date: Mon, 15 May 2023 03:14:24 +1000
FWIW, part of testing attempts to read from a failing disk i had written
code to do similar with kernel options. i vaguely meant to try to convert
it to using some sort of config flags so it's per-device capable, not all
devices, that's why i didn't commit or send for review yet.
you'd use it for your need as:
options SATA_FORCE_SPD=SControl_SPD_G1
.mrg.
Index: sys/dev/ata/sata_subr.c
===================================================================
RCS file: /cvsroot/src/sys/dev/ata/sata_subr.c,v
retrieving revision 1.24
diff -p -u -r1.24 sata_subr.c
--- sys/dev/ata/sata_subr.c 21 Jun 2018 21:52:15 -0000 1.24
+++ sys/dev/ata/sata_subr.c 14 May 2023 17:12:15 -0000
@@ -93,7 +93,11 @@ sata_reset_interface(struct ata_channel
* the device. It doesn't hurt for other devices.
*/
bus_space_write_4(sata_t, scontrol_r, 0, 0);
+#ifdef SATA_FORCE_SPD
+ scontrol = SControl_IPM_NONE | SATA_FORCE_SPD | SControl_DET_INIT;
+#else
scontrol = SControl_IPM_NONE | SControl_SPD_ANY | SControl_DET_INIT;
+#endif
bus_space_write_4(sata_t, scontrol_r, 0, scontrol);
ata_delay(chp, 50, "sataup", flags);
|
|