Linux 2.5.69-dj1

Mostly a real big set (~200KB) of AGP changes that I'd like people to give a testing, but a bunch of other bits and pieces too.

ftp://ftp.kernel.org/pub/linux/kernel/people/davej/patches/2.5/2.5.69/

Dave

Changes since 2.5.60-dj2:
- -dj leftovers
Drop lots of unneeded/broken bits.
- Lots of other leftovers from older -dj patches
This stuff needs sorting, and some more bits throwing out. There's lots of 2.4 forward port stuff here, and still lots more to come.
- Latest from Linus' bk tree.
(Just there to help me stay in sync)
- Latest bits from the cpufreq tree.
[email protected], 2003-05-01 00:09:19+01:00, [email protected]
[CPUFREQ] Remove not needed ;'s from macro definitions.

[email protected], 2003-04-28 18:13:40+01:00, [email protected]
[CPUFREQ] Acer Aspire's have broken PST tables in one BIOS rev. DMI blacklist it.

[email protected], 2003-04-24 13:14:23+01:00, [email protected]
[CPUFREQ] Fix powernow-k7 hang.
Some laptops don't like jumping from (for eg) 800MHz->1200MHz, but will happily go from 800->1000->1200. The problem is caused by us changing the FID and the VID at the same time. The spec says we have to change them seperately. This actually buys us something extra anyway, as we can now set the FID lazily as well as the VID.

- Latest pull of agpgart tree, with following csets..

[email protected], 2003-05-10 14:35:12+01:00, [email protected]
[AGPGART] Make the agp 3.5 use the agp3 code for enabling, leaving just the isoch stuff in isoch.c
[email protected], 2003-05-10 10:03:51+01:00, [email protected]
[AGPGART] CodingStyle nitpicks for isoch.c

[email protected], 2003-05-10 09:41:28+01:00, [email protected]
[AGPGART] Use symbolic defines for isoch registers in isoch code.

[email protected], 2003-05-10 09:25:16+01:00, [email protected]
[AGPGART] Add missing #defines from last checkin.

[email protected], 2003-05-10 09:22:49+01:00, [email protected]
[AGPGART] Make sure we don't poke reserved bits when enabling agp v3

[email protected], 2003-05-10 09:13:50+01:00, [email protected]
[AGPGART] Add proper AGP3 initialisation routine.

[email protected], 2003-05-10 08:15:16+01:00, [email protected]
[AGPGART] death of generic-3.0.c = folded into generic.c

[email protected], 2003-05-09 12:31:04+01:00, [email protected]
[AGPGART] Remove duplicate copying of ->chipset in agp_copy_info()

[email protected], 2003-05-06 19:24:37+01:00, [email protected]
[AGPGART] Remove unneeded exports.
These functions should only be called indirectly from agp_generic_enable()

[email protected], 2003-05-06 19:20:21+01:00, [email protected]
[AGPGART] Only enable isochronous transfers on AGP3.5 chipsets. The standard says that 3.0 chipsets don't support these extensions. Move the isoch stuff out into isoch.c leaving behind a shell for basic AGP3.0 enabling (to be written).

[email protected], 2003-05-05 21:39:31+01:00, [email protected]
[AGPGART] Work around AMD 8151 errata.
Some revisions incorrectly report they support v3.5 of the AGP spec, when they are actually only 3.0 compliant.

[email protected], 2003-05-05 21:37:30+01:00, [email protected]
[AGPGART] Store agp revision in agp_bridge struct.
There are a few places we do spec revision compliance checks, this cset generalises that function, and removes some duplicated functionality.

[email protected], 2003-05-05 15:16:17+01:00, [email protected]
[AGPGART] Skip devices with no AGP headers sooner.

[email protected], 2003-05-05 02:37:40+01:00, [email protected]
[AGPGART] Disable debugging printk's again.
With the 'AGP bug' solved, we don't need this noise for a while...

[email protected], 2003-05-03 02:21:32+01:00, [email protected]
[AGPGART] Intel I875P support.
From Matt Tolentino.

[email protected], 2003-05-01 19:35:27+01:00, [email protected]
[AGPGART] Turn on debugging printks for a while.
Lets see if we can't track down some of the stranger reports. Also fix up the macro not to printk some ': ' it doesnt need to.

[email protected], 2003-05-01 19:33:26+01:00, [email protected]
[AGPGART] missing %p in debug printk

[email protected], 2003-05-01 19:32:22+01:00, [email protected]
[AGPGART] Kill off some typedefs.
Note, I'm leaving behind the ones not in #ifdef __KERNEL__ for now, as I'm not sure just what userspace stuff that might break.

[email protected], 2003-05-01 02:59:26+01:00, [email protected]
[AGPGART] more kconfig cleanups
- Alphabetical order of items.
- Fixup the IOMMU dependancy problem (Roman Zippel)
- Word all options similarly (Add 'chipset' to some descriptions)
- Remove unused CONFIG_AGP_GART
- Fix up incomplete alpha tristate

[email protected], 2003-05-01 02:43:08+01:00, [email protected]
[AGPGART] Move debugging macros to header so they can be used in other parts of agpgart.

[email protected], 2003-05-01 02:41:16+01:00, [email protected]
[AGPGART] Fix incorrect type warning.

[email protected], 2003-05-01 01:30:58+01:00, [email protected]
[AGPGART] Remove duplicate code in i810/i830 alloc_by_type functions.

[email protected], 2003-05-01 01:29:49+01:00, [email protected]
[AGPGART] Bulletproofing. NULL ptrs after freeing them.

[email protected], 2003-04-30 23:56:19+01:00, [email protected]
[AGPGART] Add some debugging printk's. Based on Linus' earlier patch.

[email protected], 2003-04-28 02:32:41+01:00, [email protected]
[AGPGART] Fix linking error.

[email protected], 2003-04-27 21:44:07+01:00, [email protected]
[AGPGART] Shrink chipset_type enum (compile fix) Missing part of hch's last cset.

[email protected], 2003-04-27 21:15:40+01:00, [email protected]
[AGPGART] Fix Kconfig typo

[email protected], 2003-04-27 21:14:58+01:00, [email protected]
[AGPGART] proper agp_bridge_driver.
Christoph with the goods once more...

[email protected], 2003-04-27 19:52:51+01:00, [email protected]
[AGPGART] give all agpgart drivers a ->remove pci method. You guessed it, yup. from Christoph again.

[email protected], 2003-04-27 19:40:23+01:00, [email protected]
[AGPGART] don't dereference agp_bridge in generic-3.0.c Yet more from Christoph..

[email protected], 2003-04-27 19:27:24+01:00, [email protected]
[AGPGART] Add back dummy module exit to keep things happy.

[email protected], 2003-04-27 19:20:41+01:00, [email protected]
[AGPGART] Nvidia GART cleanups.
Christoph cleaned up a lot of the mess here. We're back to nearly killing off the chipset_type enum, moved the register definitions to the code that uses it, and given it a proper pci .remove function.

[email protected], 2003-04-27 19:17:31+01:00, [email protected]
[AGPGART] cleanup agp backend.c a bit
More from Christoph.

Most style nitpicks and a bit more explicitly passing struct agp_bridge_data around.

[email protected], 2003-04-27 19:03:06+01:00, [email protected]
[AGPGART] fix macros that expect agp_bridge in global scope From Christoph Hellwig

[email protected], 2003-04-25 17:43:47+01:00, davej@tetrachloride.(none)
[AGPGART] Fix kconfig dependancies.
- Don't show x86 GARTs on alpha

[email protected], 2003-04-25 17:43:41+01:00, davej@tetrachloride.(none)
[AGPGART] i855PM support from Bill Nottingham.
Mainly adding a PCI id, unfortunately, requires
renaming the i855GM PCI ids to avoid name conflict. Also renames some of the i855GM constants from i855PM to i855GM.

[email protected], 2003-04-25 17:43:35+01:00, davej@tetrachloride.(none)
[AGPGART] Remove semaphore abstraction.

[email protected], 2003-04-25 17:43:25+01:00, davej@tetrachloride.(none)
[AGPGART] Misc backend source tidy up.

[email protected], 2003-04-25 17:43:18+01:00, davej@tetrachloride.(none)
[AGPGART] Move function prototypes to headers.

[email protected], 2003-04-25 17:43:13+01:00, davej@tetrachloride.(none)
[AGPGART] kdoc'ify some of the function header comments.

[email protected], 2003-04-25 17:43:07+01:00, davej@tetrachloride.(none)
[AGPGART] Move function description comments from headers to the code they document.

[email protected], 2003-04-25 17:43:01+01:00, davej@tetrachloride.(none)
[AGPGART] EXPORT_SYMBOL cleanups. Also move the global_cache_flush routine to generic.c

[email protected], 2003-04-25 17:42:55+01:00, davej@tetrachloride.(none)
[AGPGART] Fix typo that stopped nvidia GART driver being built

[email protected], 2003-04-25 17:42:48+01:00, davej@tetrachloride.(none)
[AGPGART] Fall back to non-isochronous xfers if setting up isochronous xfers fails.

[email protected], 2003-04-25 17:42:42+01:00, davej@tetrachloride.(none)
[AGPGART] Convert several functions to return void.
They only ever returned a single value.

[email protected], 2003-04-25 17:42:35+01:00, davej@tetrachloride.(none)
[AGPGART] use symbols instead of hardcoded values in generic-3.0 Lots more work to do here.

[email protected], 2003-04-25 17:42:29+01:00, davej@tetrachloride.(none)
[AGPGART] Don't configure agp bridges more than once if there is >1 of them.

[email protected], 2003-04-25 17:42:22+01:00, davej@tetrachloride.(none)
[AGPGART] Add more defines to kill off hardcoded values

[email protected], 2003-04-25 17:42:16+01:00, davej@tetrachloride.(none)
[AGPGART] Add symbolic constants for AGP mode setting.
bye bye icky hardcoded values.

[email protected], 2003-04-25 17:42:10+01:00, davej@tetrachloride.(none)
[AGPGART] Remove unneeded settings of bridge->type.
It's now done in the static structs.

[email protected], 2003-04-25 17:42:03+01:00, davej@tetrachloride.(none)
[AGPGART] Makefile cleanups.
- the makefile is not the right place to describe the driver
- remove some junk

[email protected], 2003-04-25 17:41:58+01:00, davej@tetrachloride.(none)
[AGPGART] Merge NVIDIA nForce / nForce2 AGP driver.
Based upon code written by NVIDIA for agpgart 2.4, forward ported and cleaned up slightly by me. This still needs work, and is untested.

[email protected], 2003-04-25 17:41:51+01:00, davej@tetrachloride.(none)
[AGPGART] Replace enum users with own methods.
By introducing a few extra functions, we can kill off a few extra members of the enum. More work from Christoph Hellwig.

[email protected], 2003-04-25 17:41:45+01:00, davej@tetrachloride.(none)
[AGPGART] More setup routine -> static struct conversions. Again from Christoph Hellwig.

[email protected], 2003-04-25 17:41:38+01:00, davej@tetrachloride.(none)
[AGPGART] Remove unneeded enums from AMD k7 gart driver

[email protected], 2003-04-25 17:41:32+01:00, davej@tetrachloride.(none)
[AGPGART] Remove useless enums from serverworks gart driver

[email protected], 2003-04-25 17:41:26+01:00, davej@tetrachloride.(none)
[AGPGART] Fix typo in via-agp. s/PM400/P4M400/

[email protected], 2003-04-25 17:41:19+01:00, davej@tetrachloride.(none)
[AGPGART] Remove stale comment

[email protected], 2003-04-25 17:41:13+01:00, davej@tetrachloride.(none)
[AGPGART] Remove unused ALi enums.

[email protected], 2003-04-25 17:41:06+01:00, davej@tetrachloride.(none)
[AGPGART] Remove unneeded enums from intel gart driver.

[email protected], 2003-04-25 17:40:59+01:00, davej@tetrachloride.(none)
[AGPGART] intel agp init cleanups.
From Christoph Hellwig.
(1) Kill the _setup functions - most of it can be nice, static structs - the few remainders are handled better elsewhere
(2) _one_ big switch in ->probe assigning these tables.

Almost 200 lOC gone and it's even readable :)

[email protected], 2003-04-25 17:40:50+01:00, davej@tetrachloride.(none)
[AGPGART] Hammer GART can use generic enable routines now.

<-- snip -->

...
gcc -Wp,-MD,drivers/net/.tlan.o.d -D__KERNEL__ -Iinclude -Wall
-Wstrict-prototypes -Wno-trigraphs -O2 -fno-strict-aliasing -fno-common -pipe -mpreferred-stack-boundary=2 -march=k6 -Iinclude/asm-i386/mach-default -nostdinc -iwithprefix include -DKBUILD_BASENAME=tlan -DKBUILD_MODNAME=tlan -c -o drivers/net/tlan.o drivers/net/tlan.c
drivers/net/tlan.c:1539:21: warning: extra tokens at end of #ifdef directive drivers/net/tlan.c:1540:2: #error "Not 64bit clean"
make[2]: *** [drivers/net/tlan.o] Error 1

<-- snip -->

This is a bug only present in -dj, the fix is simple:

@@ -1536,7 +1536,7 @@
 				t = (void *) skb_put( new_skb, TLAN_MAX_FRAME_SIZE );
 				head_list->buffer[0].address = pci_map_single(priv->pciDev, new_skb->data, TLAN_MAX_FRAME_SIZE, PCI_DMA_FROMDEVICE);
 				head_list->buffer[8].address = (u32) t;
-#ifdef BITS_PER_LONG==64
+#if BITS_PER_LONG == 64
 #error "Not 64bit clean"
 #endif				
 				head_list->buffer[9].address = (u32) new_skb;

cu
Adrian

       "Is there not promise of rain?" Ling Tan asked suddenly out
        of the darkness. There had been need of rain for many days.
       "Only a promise," Lao Er said.
                                       Pearl S. Buck - Dragon Seed

Hi Dave,

the changes to sound/isa/dt019x.c (ALSA) in 2.5.69-dj1 are bogus, it seems this is the reversion of a adaption to the PNP changes and it results in a compile error in -dj.

cu
Adrian

       "Is there not promise of rain?" Ling Tan asked suddenly out
        of the darkness. There had been need of rain for many days.
       "Only a promise," Lao Er said.
                                       Pearl S. Buck - Dragon Seed

-dj makes agp_init static resulting in the following error at the final linking:

<-- snip -->

...
386/oprofile/built-in.o net/built-in.o --end-group -o .tmp_vmlinux1 drivers/built-in.o(.init.text+0x6b584): In function `i810fb_init':
: undefined reference to `agp_init'
make: *** [.tmp_vmlinux1] Error 1

<-- snip -->

The following patch fixes it:

@@ -307,7 +307,7 @@
 EXPORT_SYMBOL_GPL(agp_remove_bridge);
 
 
-static int __init agp_init(void)
+int __init agp_init(void)
 {
 	printk(KERN_INFO "Linux agpgart interface v%d.%d (c) Dave Jones\n",
 	       AGPGART_VERSION_MAJOR, AGPGART_VERSION_MINOR);

cu
Adrian

       "Is there not promise of rain?" Ling Tan asked suddenly out
        of the darkness. There had been need of rain for many days.
       "Only a promise," Lao Er said.
                                       Pearl S. Buck - Dragon Seed

duhh, the 810 framebuffer needs it early. I forgot about that. Will apply patch, and add a comment. Thanks.

Dave

no, it doesn't need the agp banner printk early :) Fix i810fb instead.

Ahh crap, this may bring up another problem. The agpgart _has_ to be initialised before the i810fb code, otherwise it won't work. Now that agp_init() doesn't do anything useful, we're relying on link order. Whether we get that right or not right now depends.. Needs to be tested by someone who actually uses i810fb to be sure.

Volunteers? (Just chop out the agp_init call in drivers/video/i810/i810_main.c), oh and I'm only interested in feedback from either users of 2.5.69-dj1, or 2.5 bitkeeper tree, (Linus seems to have taken the first round of AGP updates in the last few hours).

Dave