Most recent kernel where this bug did not occur: 2.6.12-1.1447_FC4smp (which is a 2.6.12 with acpi patch 20050729) Distribution: Fedora Core 4 Hardware Environment: Acer Aspire T630 Software Environment: gcc-4.0.1-4.fc4 Problem Description: The previous kernel 2.6.12-1.1398_FC4smp worked with ACPI loaded. It has the acpi patch 20050309. I tried to recompile the kernel with the acpi patch 20050408 and it worked too. But now with the last acpi patch, I've got a failure while initializing ACPI. Consequence ACPI is disabled, and some other problems occur, like sata_uli modules which hangs; more info here: https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=165103 I recompiled with ACPI debug; here's a bit of the output of dmesg: BIOS-provided physical RAM map: BIOS-e820: 0000000000000000 - 000000000009fc00 (usable) BIOS-e820: 000000000009fc00 - 00000000000a0000 (reserved) BIOS-e820: 00000000000f0000 - 0000000000100000 (reserved) BIOS-e820: 0000000000100000 - 000000001fff0000 (usable) BIOS-e820: 000000001fff0000 - 000000001fff3000 (ACPI NVS) BIOS-e820: 000000001fff3000 - 0000000020000000 (ACPI data) BIOS-e820: 00000000e0000000 - 00000000f0000000 (reserved) BIOS-e820: 00000000fec00000 - 0000000100000000 (reserved) .... ACPI: RSDP (v000 AWARD ) @ 0x000f7110 ACPI: RSDT (v001 AWARD AWRDACPI 0x42302e31 AWRD 0x01010101) @ 0x1fff3040 ACPI: FADT (v001 AWARD AWRDACPI 0x42302e31 AWRD 0x01010101) @ 0x1fff30c0 ACPI: MCFG (v001 AWARD AWRDACPI 0x42302e31 AWRD 0x01010101) @ 0x1fff7280 ACPI: MADT (v001 AWARD AWRDACPI 0x42302e31 AWRD 0x01010101) @ 0x1fff71c0 ACPI: DSDT (v001 AWARD AWRDACPI 0x00001000 MSFT 0x0100000c) @ 0x00000000 ACPI: PM-Timer IO Port: 0x1008 Allocating PCI resources starting at 20000000 (gap: 20000000:c0000000) Built 1 zonelists .... CPU: Intel(R) Pentium(R) 4 CPU 3.20GHz stepping 01 Enabling fast FPU save and restore... done. Enabling unmasked SIMD FPU exception support... done. Checking 'hlt' instruction... OK. tbxface-0109 [02] load_tables : ACPI Tables successfully acquired dswstate-0233 [13] ds_result_pop_from_bot: No result objects! State=dfeb0828 dsutils-0641 [13] ds_create_operand : Missing or null operand, AE_AML_NO_RETURN_VALUE dsutils-0734 [12] ds_create_operands : While creating Arg 1 - AE_AML_NO_RETURN_VALUE tbxface-0115: *** Error: acpi_load_tables: Could not load namespace: AE_AML_NO_RETURN_VALUE tbxface-0123: *** Error: acpi_load_tables: Could not load tables: AE_AML_NO_RETURN_VALUE ACPI: Unable to load the System Description Tables checking if image is initramfs... it is Freeing initrd memory: 1195k freed NET: Registered protocol family 16 PCI: PCI BIOS revision 3.00 entry at 0xfaf50, last bus=2 PCI: Using MMCONFIG ACPI: Subsystem revision 20050729 ACPI: Interpreter disabled. Linux Plug and Play Support v0.97 (c) Adam Belay pnp: PnP ACPI: disabled
please attach the output from acpidump available in pmtools here: http://ftp.kernel.org/pub/linux/kernel/people/lenb/acpi/utils/
Created attachment 5841 [details] output of acpidump run with kernel 2.6.12-1.1398_FC4smp
This DSDT seems to have the "executable code at module level" which is not currently supported by ACPICA, and results are unpredictable. Store (Buffer (0x03) {}, Local0) CreateByteField (Local0, 0x00, BUF1) CreateByteField (Local0, 0x01, BUF2) CreateByteField (Local0, 0x02, DMAM) Versions of ACPICA beyond 20050513 cannot disassemble the DSDT, this is a problem also. We are still investigating the solution for both problems.
Problems with the support for module-level executable code forced us to disable the code. However, some of the support remained active, causing these problems. Fixed for the next ACPICA release. Also fixes the disassembler issue.
Fixed in ACPICA version 20050902
Shipped in 2.6.13-git9. Closing.
I've tested with kernel 2.6.13-git11 with SMP feature. It works, but of course there are less entries in /proc/acpi compared to my current kernel 2.6.12-1.1398_FC4smp (alarm, sleep, ...).