Date: Fri, 3 Sep 2004 11:40:49 -0700 (PDT) From: Don Lewis <dl@catspoiler.org> To: martines@rochester.rr.com Cc: scsi@FreeBSD.org Subject: smartmontools on FreeBSD 4.x - compile error + ahc card dump Message-ID: <200409031840.i83IenAj029786@gw.catspoiler.org>
next in thread | raw e-mail | index | archive | help
The latest version of smartmontools is broken on FreeBSD 4.x (and older
versions of 5.x that don't have the latest version of <inttypes.h>). I
can get the code to compile with the patch below, but then I get runtime
kernel errors. Smartmontools-5.30 seems to work fine on FreeBSD 4.x.
The only suspicious change that I've seen so far is that
smartmontools-5.32 passes a non-zero timeout to do_scsi_cmnd_io(), which
passes it to cam_fill_csio() and cam_send_ccb(), whereas 5.30 passes
zero in most cases. Suspicion confirmed ... if I force the timeout to
zero in do_scsi_cmnd_io(), smartctl -A works properly.
ahc0: Recovery Initiated
>>>>>>>>>>>>>>>>>> Dump Card State Begins <<<<<<<<<<<<<<<<<
ahc0: Dumping Card State while idle, at SEQADDR 0x8
Card was paused
ACCUM = 0x4, SINDEX = 0x64, DINDEX = 0x65, ARG_2 = 0x0
HCNT = 0x0 SCBPTR = 0x19
SCSIPHASE[0x8]:(MSG_IN_PHASE) SCSISIGI[0xe6]:(REQI|BSYI|MSGI|IOI|CDI)
ERROR[0x0] SCSIBUSL[0x80] LASTPHASE[0x1]:(P_BUSFREE)
SCSISEQ[0x12]:(ENAUTOATNP|ENRSELI) SBLKCTL[0xa]:(SELWIDE|SELBUSB)
SCSIRATE[0x0] SEQCTL[0x10]:(FASTMODE) SEQ_FLAGS[0xc0]:(NO_CDB_SENT|NOT_IDENTIFIED)
SSTAT0[0x20]:(SELDI) SSTAT1[0x19]:(REQINIT|BUSFREE|PHASEMIS)
SSTAT2[0x0] SSTAT3[0x0] SIMODE0[0x8]:(ENSWRAP) SIMODE1[0xac]:(ENSCSIPERR|ENBUSFREE|ENSCSIRST|ENSELTIMO)
SXFRCTL0[0x80]:(DFON) DFCNTRL[0x0]
DFSTATUS[0x89]:(FIFOEMP|HDONE|PRELOAD_AVAIL)
STACK: 0x0 0x164 0x179 0x3
SCB count = 70
Kernel NEXTQSCB = 25
Card NEXTQSCB = 25
QINFIFO entries:
Waiting Queue entries:
Disconnected Queue entries: 25:0
QOUTFIFO entries:
Sequencer Free SCB List: 18 14 11 7 26 3 13 22 5 8 19 15 6 1 17 28 27 24 30
16 10 23 0 4 31 2 9 29 21 20 12
Sequencer SCB Info:
0 SCB_CONTROL[0xe0]:(TAG_ENB|DISCENB|TARGET_SCB) SCB_SCSIID[0x7]
SCB_LUN[0x0] SCB_TAG[0xff]
1 SCB_CONTROL[0xe0]:(TAG_ENB|DISCENB|TARGET_SCB) SCB_SCSIID[0x7]
SCB_LUN[0x0] SCB_TAG[0xff]
2 SCB_CONTROL[0xe0]:(TAG_ENB|DISCENB|TARGET_SCB) SCB_SCSIID[0x7]
SCB_LUN[0x0] SCB_TAG[0xff]
3 SCB_CONTROL[0xe0]:(TAG_ENB|DISCENB|TARGET_SCB) SCB_SCSIID[0x7]
SCB_LUN[0x0] SCB_TAG[0xff]
4 SCB_CONTROL[0xe0]:(TAG_ENB|DISCENB|TARGET_SCB) SCB_SCSIID[0x7]
SCB_LUN[0x0] SCB_TAG[0xff]
5 SCB_CONTROL[0xe0]:(TAG_ENB|DISCENB|TARGET_SCB) SCB_SCSIID[0x7]
SCB_LUN[0x0] SCB_TAG[0xff]
6 SCB_CONTROL[0xe0]:(TAG_ENB|DISCENB|TARGET_SCB) SCB_SCSIID[0x7]
SCB_LUN[0x0] SCB_TAG[0xff]
7 SCB_CONTROL[0xe0]:(TAG_ENB|DISCENB|TARGET_SCB) SCB_SCSIID[0x7]
SCB_LUN[0x0] SCB_TAG[0xff]
8 SCB_CONTROL[0xe0]:(TAG_ENB|DISCENB|TARGET_SCB) SCB_SCSIID[0x7]
SCB_LUN[0x0] SCB_TAG[0xff]
9 SCB_CONTROL[0xe0]:(TAG_ENB|DISCENB|TARGET_SCB) SCB_SCSIID[0x7]
SCB_LUN[0x0] SCB_TAG[0xff]
10 SCB_CONTROL[0xe0]:(TAG_ENB|DISCENB|TARGET_SCB) SCB_SCSIID[0x7]
SCB_LUN[0x0] SCB_TAG[0xff]
11 SCB_CONTROL[0xe0]:(TAG_ENB|DISCENB|TARGET_SCB) SCB_SCSIID[0x7]
SCB_LUN[0x0] SCB_TAG[0xff]
12 SCB_CONTROL[0xe0]:(TAG_ENB|DISCENB|TARGET_SCB) SCB_SCSIID[0x7]
------ Forwarded message ------
From: Don Lewis <truckman@FreeBSD.org>
Subject: Re: cvs commit: ports/sysutils/smartmontools Makefile
Date: Fri, 3 Sep 2004 10:29:39 -0700 (PDT)
To: stolz@i2.informatik.rwth-aachen.de
On 3 Sep, Volker Stolz wrote:
> Am 03. Sep 2004 um 11:17 CEST schrieb Don Lewis:
>> On 3 Sep, Volker Stolz wrote:
>> > vs 2004-09-03 08:32:28 UTC
>> >
>> > FreeBSD ports repository
>> >
>> > Modified files:
>> > sysutils/smartmontools Makefile
>> > Log:
>> > Mark as IGNORE for OSVERSION < 501105: Needs ATAng
>>
>> Not quite true. It works fine with SCSI.
>
> But the port doesn't compile. Do you have any patches or could you
> tell me where to start looking? Is there a switch for disabling the
> breaking part of the build?>
> Cheers,
> Volker
The following patch fixes the compile problem on 4-STABLE. The FreeBSD
version is the closest to when <inttypes.h> was changed to match the
standard. I get a bunch of SCSI errors when I try to run it though.
--- int64.h.orig Mon Mar 15 11:47:22 2004
+++ int64.h Fri Sep 3 10:13:56 2004
@@ -35,6 +35,9 @@
#ifdef HAVE_STDINT_H
#include <stdint.h>
#else
+#if __FreeBSD_version <= 500042
+#include <inttypes.h>
+#else
#if defined(_WIN32) && defined(_MSC_VER)
// for MSVC 6.0
typedef __int64 int64_t;
@@ -44,11 +47,12 @@
typedef long long int64_t;
typedef unsigned long long uint64_t;
#endif // _WIN32 && _MSC_VER
+#endif // __FreeBSD_version <= 500042
#endif // HAVE_STDINT_H
#endif // HAVE_SSYS_INT_TYPES_H
// 64 bit integer format strings
-#ifdef HAVE_INTTYPES_H
+#if defined(HAVE_INTTYPES_H) && __FreeBSD_version > 500042
#include <inttypes.h>
#else
#if defined(_WIN32) && defined(_MSC_VER)
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?200409031840.i83IenAj029786>