From owner-freebsd-current@FreeBSD.ORG Tue May 6 14:28:21 2003 Return-Path: Delivered-To: freebsd-current@freebsd.org Received: from mx1.FreeBSD.org (mx1.freebsd.org [216.136.204.125]) by hub.freebsd.org (Postfix) with ESMTP id 0521937B405 for ; Tue, 6 May 2003 14:28:21 -0700 (PDT) Received: from rootlabs.com (root.org [67.118.192.226]) by mx1.FreeBSD.org (Postfix) with SMTP id 885EB43F3F for ; Tue, 6 May 2003 14:28:20 -0700 (PDT) (envelope-from nate@rootlabs.com) Received: (qmail 50172 invoked by uid 1000); 6 May 2003 21:28:21 -0000 Date: Tue, 6 May 2003 14:28:21 -0700 (PDT) From: Nate Lawson To: acpi-jp@jp.freebsd.org Message-ID: <20030506140539.F18533@root.org> MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII cc: current@freebsd.org Subject: Outstanding ACPI issues for 5.1-RELEASE X-BeenThere: freebsd-current@freebsd.org X-Mailman-Version: 2.1.1 Precedence: list List-Id: Discussions about the use of FreeBSD-current List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , X-List-Received-Date: Tue, 06 May 2003 21:28:21 -0000 The current state of FreeBSD ACPI is that the Intel 0228 dist was imported along with a one-line fix to hwregs.c that solves power-off problems. Here is a list of the current problems. We would like to fix as many as possible before the release. If anyone has time to tackle these issues, please let me know. The reporters listed below should open a PR and send me the number as I do not want to replicate GNATS in email. -------------------------------------------------------------- Problem: acpi_thermal causes a panic Reporter: ru@freebsd.org Information: This looks like referencing a freed or uninitialized data member. The below was produces with the latest 0228 ACPI diff in-tree, and acpi device compiled directly into a kernel. Loading it as a module gave me some lines offset in the source file, and the faulting address was 0xdeadc0de. The box.* files and "pciconf -l" output are available at: http://people.freebsd.org/~ru/acpi_thermal.panic.bz2 : #11 0xc02d5968 in calltrap () at {standard input}:96 : #12 0xc013c0fd in AcpiGetHandle (Parent=0xbffff, Pathname=0xc030d14e "_PS0", : RetHandle=0xc8761c2c) at /usr/src/sys/contrib/dev/acpica/nsxfname.c:177 : #13 0xc0153f83 in acpi_pwr_switch_consumer (consumer=0x0, state=0) : at /usr/src/sys/dev/acpica/acpi_powerres.c:359 : #14 0xc0155818 in acpi_tz_switch_cooler_on (obj=0xc189c9d0, arg=0xc1878300) : at /usr/src/sys/dev/acpica/acpi_thermal.c:588 : #15 0xc014ba8d in acpi_ForeachPackageObject (pkg=0xc1878300, : func=0xc01557e0 , arg=0xc1878300) : at /usr/src/sys/dev/acpica/acpi.c:1186 : #16 0xc01555fb in acpi_tz_monitor (sc=0xc189c9d0) : at /usr/src/sys/dev/acpica/acpi_thermal.c:462 : #17 0xc0155b66 in acpi_tz_timeout (sc=0xc1878300) : at /usr/src/sys/dev/acpica/acpi_thermal.c:741 : #18 0xc0155d46 in acpi_tz_thread (arg=0x0) : at /usr/src/sys/dev/acpica/acpi_thermal.c:821 -------------------------------------------------------------- Problem: Uninstalling multiple SSDT tables Reporter: Marcel The HP rx2600 machines (ia64) have multiple SSDT tables. These tables get nicely linked together. The first table descriptor is static, the following table descriptors are allocated. When shutting down the machine, these tables (and descriptors) will get uninstalled. This however contains a bug: the ListHead is removed first and since (head->next != head->prev) when more than one instance of a table exists, it is not identified as the list head and thus is being deallocated. A tested patch is attached first. A minimal (untested) patch is attached second. I like the second for it's simplicity. Both patches try to fix the same problem: deallocation (freeing) memory that was not previously allocated. [Patch available, email me for it] -------------------------------------------------------------- Problem: UtAllocate zero bytes warnings Reporter: Warner Losh Additional reporters: Thorsten Greiner ACPI-0293: *** Warning: Buffer created with zero length in AML -0166: *** Error: UtAllocate: Attempt to allocate zero bytes DSDT and other info is here: http://www.stethojupi.de/acpi.tar.bz2 -------------------------------------------------------------- Problem: AML_BUFFER_LIMIT errors on boot Reporter: rwatson@freebsd.org Information: Tons of errors printed at boot. ACPI-1287: *** Error: Method execution failed [\\_SB_.PSRS] (Node 0xc0b77ca0), AE_AML_BUFFER_LIMIT ACPI-1287: *** Error: Method execution failed [\\_SB_.PIIF._SRS] (Node 0xc188cc80), AE_AML_BUFFER_LIMIT ACPI-1287: *** Error: Method execution failed [\\_SB_.PSRS] (Node 0xc0b77ca0), AE_AML_BUFFER_LIMIT ACPI-1287: *** Error: Method execution failed [\\_SB_.PIIF._SRS] (Node 0xc188cc80), AE_AML_BUFFER_LIMIT ACPI-1287: *** Error: Method execution failed [\\_SB_.PSRS] (Node 0xc0b77ca0), AE_AML_BUFFER_LIMIT ACPI-1287: *** Error: Method execution failed [\\_SB_.PIIF._SRS] (Node 0xc188cc80), AE_AML_BUFFER_LIMIT ACPI-1287: *** Error: Method execution failed [\\_SB_.PSRS] (Node 0xc0b77ca0), AE_AML_BUFFER_LIMIT ACPI-1287: *** Error: Method execution failed [\\_SB_.PIIF._SRS] (Node 0xc188cc80), AE_AML_BUFFER_LIMIT ACPI-1287: *** Error: Method execution failed [\\_SB_.PSRS] (Node 0xc0b77ca0), AE_AML_BUFFER_LIMIT