If HW IOMMU initialization fails (Intel VT-d often does this,
typically due to BIOS bugs), we fall back to nommu. It doesn't
work for the majority since nowadays we have more than 4GB
memory so we must use swiotlb instead of nommu.
The problem is that it's too late to initialize swiotlb when HW
IOMMU initialization fails. We need to allocate swiotlb memory
earlier from bootmem allocator. Chris explained the issue in
detail:
http://marc.info/?l=linux-kernel&m=125657444317079&w=2
The current x86 IOMMU initialization sequence is too complicated
and handling the above issue makes it more hacky.
This patch changes x86 IOMMU initialization sequence to handle
the above issue cleanly.
The new x86 IOMMU initialization sequence are:
1. we initialize the swiotlb (and setting swiotlb to 1) in the case
of (max_pfn > MAX_DMA32_PFN && !no_iommu). dma_ops is set to
swiotlb_dma_ops or nommu_dma_ops. if swiotlb usage is forced by
the boot option, we finish here.
2. we call the detection functions of all the IOMMUs
3. the detection function sets x86_init.iommu.iommu_init to the
IOMMU initialization function (so we can avoid calling the
initialization functions of all the IOMMUs needlessly).
4. if the IOMMU initialization function doesn't need to swiotlb
then sets swiotlb to zero (e.g. the initialization is
sucessful).
5. if we find that swiotlb is set to zero, we free swiotlb
resource.
Signed-off-by: FUJITA Tomonori <fujita.tomonori@lab.ntt.co.jp>
Cc: chrisw@sous-sol.org
Cc: dwmw2@infradead.org
Cc: joerg.roedel@amd.com
Cc: muli@il.ibm.com
LKML-Reference: <1257849980-22640-10-git-send-email-fujita.tomonori@lab.ntt.co.jp>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
This changes gart_iommu_hole_init() to set gart_iommu_init() to
iommu_init hook if gart_iommu_hole_init() finds the GART IOMMU.
We can kill the code to check if we found the IOMMU in
gart_iommu_init() since gart_iommu_hole_init() sets
gart_iommu_init() only when it found the IOMMU.
Signed-off-by: FUJITA Tomonori <fujita.tomonori@lab.ntt.co.jp>
Cc: chrisw@sous-sol.org
Cc: dwmw2@infradead.org
Cc: joerg.roedel@amd.com
Cc: muli@il.ibm.com
LKML-Reference: <1257849980-22640-4-git-send-email-fujita.tomonori@lab.ntt.co.jp>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
This block is allocated with alloc_bootmem() and scanned by kmemleak but
the kernel direct mapping may no longer exist. This patch tells kmemleak
to ignore this memory hole. The dma32_bootmem_ptr in
dma32_reserve_bootmem() is also ignored.
Signed-off-by: Catalin Marinas <catalin.marinas@arm.com>
Acked-by: Ingo Molnar <mingo@elte.hu>
Impact: update comment
Clarify that too small aperture is valid reason for this code.
Signed-off-by: Pavel Machek <pavel@suse.cz>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
The number of BIOSes that have an option to enable the IOMMU, or fix
anything about its configuration, is vanishingly small. There's no good
reason to punish quiet boot for this.
Signed-off-by: Adam Jackson <ajax@redhat.com>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
gart.h has only GART-specific stuff. Only GART code needs it. Other
IOMMU stuff should include iommu.h instead of gart.h.
Signed-off-by: FUJITA Tomonori <fujita.tomonori@lab.ntt.co.jp>
Acked-by: Muli Ben-Yehuda <muli@il.ibm.com>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
rename update_memory_range to e820_update_range
rename add_memory_region to e820_add_region
to make it more clear that they are about e820 map operations.
Signed-off-by: Yinghai Lu <yhlu.kernel@gmail.com>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
Kevin Winchester reported a GART related direct rendering failure against
linux-next-20080611, which shows up via these log entries:
PCI: Using ACPI for IRQ routing
PCI: Cannot allocate resource region 0 of device 0000:00:00.0
agpgart: Detected AGP bridge 0
agpgart: Aperture conflicts with PCI mapping.
agpgart: Aperture from AGP @ e0000000 size 128 MB
agpgart: Aperture conflicts with PCI mapping.
agpgart: No usable aperture found.
agpgart: Consider rebooting with iommu=memaper=2 to get a good aperture.
instead of the expected:
PCI: Using ACPI for IRQ routing
agpgart: Detected AGP bridge 0
agpgart: Aperture from AGP @ e0000000 size 128 MB
Kevin bisected it down to this change in tip/x86/gart:
"x86: checking aperture size order".
agp check is using request_mem_region(), and could fail if e820 is reserved...
change it back to e820_any_mapped().
Reported-and-bisected-by: "Kevin Winchester" <kjwinchester@gmail.com>
Signed-off-by: Yinghai Lu <yhlu.kernel@gmail.com>
Tested-by: Kevin Winchester <kjwinchester@gmail.com>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
If GART IOMMU is used on an AMD64 system, the northbridge registers
related to it should be restored during resume so that memory is not
corrupted. Make gart_resume() handle that as appropriate.
Ref. http://lkml.org/lkml/2008/5/25/96 and the following thread.
Signed-off-by: Rafael J. Wysocki <rjw@sisk.pl>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
If
fix == 0, aper_enabled == 1, gart_fix_e820 == 0
if (!fix && !aper_enabled)
return;
if (gart_fix_e820 && !fix && aper_enabled) {
if (e820_any_mapped(aper_base, aper_base + aper_size,
E820_RAM)) {
/* reserve it, so we can reuse it in second kernel */
printk(KERN_INFO "update e820 for GART\n");
add_memory_region(aper_base, aper_size, E820_RESERVED);
update_e820();
}
return;
}
/* different nodes have different setting, disable them all atfirst*/
we'll fall back here and disable all the settings, even when they were
all consistent.
What about this? (I hope it compiles...)
Signed-off-by: Pavel Machek <pavel@suse.cz>
Cc: Dave Jones <davej@codemonkey.org.uk>
Cc: Andi Kleen <andi@firstfloor.org>
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Hi!
void __init early_gart_iommu_check(void)
contains
for (num = 24; num < 32; num++) {
if (!early_is_k8_nb(read_pci_config(0, num, 3, 0x00)))
continue;
loop, with very similar loop duplicated in
void __init gart_iommu_hole_init(void)
. First copy of a loop seems to be buggy, too. It uses 0 as a "nothing
set" value, which may actually bite us in last_aper_enabled case
(because it may be often zero).
(Beware, it is hard to test this patch, because this code has about
2^8 different code paths, depending on hardware and cmdline settings).
Plus, the second loop does not check for consistency of
aper_enabled. Should it?
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Some small cleanups for aperture_64.c; they should not really change
any code.
Signed-off-by: Pavel Machek <pavel@suse.cz>
Cc: Dave Jones <davej@codemonkey.org.uk>
Cc: Andi Kleen <andi@firstfloor.org>
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
1. use symbolic register names where appropriate.
2. num to bus or slot changing
3. handle for new opteron for bus other than 0
Signed-off-by: Yinghai Lu <yhlu.kernel@gmail.com>
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
because we try to reserve dma32 early, so we have chance to get aperture
from 64M.
with some sequence aperture allocated from RAM, could become E820_RESERVED.
and then if doing a kexec with a big kernel that uncompressed size is above
64M we could have a range conflict with still using gart.
So allocate gart aperture from 512M instead.
Also change the fallback_aper_order to 5, because we don't have chance to get
2G or 4G aperture.
Signed-off-by: Yinghai Lu <yhlu.kernel@gmail.com>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
some systems are using 32M for gart and agp when memory is less than 4G.
Kernel will reject and try to allcate another 64M that is not needed,
and we will waste 64M of perfectly good RAM.
this patch adds a workaround by checking aper_base/order between NB and
agp bridge. If they are the same, and memory size is less than 4G, it
will allow it.
Signed-off-by: Yinghai Lu <yhlu.kernel@gmail.com>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
while looking at Rafael J. Wysocki's system boot log,
I found a funny printout:
Node 0: aperture @ de000000 size 32 MB
Aperture too small (32 MB)
AGP bridge at 00:04:00
Aperture from AGP @ de000000 size 4096 MB (APSIZE 0)
Aperture too small (0 MB)
Your BIOS doesn't leave a aperture memory hole
Please enable the IOMMU option in the BIOS setup
This costs you 64 MB of RAM
Mapping aperture over 65536 KB of RAM @ 4000000
...
agpgart: Detected AGP bridge 20
agpgart: Aperture pointing to RAM
agpgart: Aperture from AGP @ de000000 size 4096 MB
agpgart: Aperture too small (0 MB)
agpgart: No usable aperture found.
agpgart: Consider rebooting with iommu=memaper=2 to get a good aperture.
it means BIOS allocated the correct gart on the NB and AGP bridge, but
because a bug in the silicon (the agp bridge reports the wrong order,
it wants 4G instead) the kernel will reject that allocation.
Also, because the size is only 32MB, and we try to get another 64M for gart,
late fix_northbridge can not revert that change because it still reads
the wrong size from agp bridge.
So try to double check the order value from the agp bridge, before calling
aperture_valid().
[ mingo@elte.hu: 32-bit fix. ]
Signed-off-by: Yinghai Lu <yhlu.kernel@gmail.com>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Initializing to zero is generally bad idea, I hope it is right for
__init data, too.
Signed-off-by: Pavel Machek <pavel@suse.cz>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
aperture_64.c takes a piece of memory and makes it into iommu
window... but such window may not be saved by swsusp -- that leads to
oops during hibernation.
Signed-off-by: Pavel Machek <pavel@suse.cz>
Acked-by: "Rafael J. Wysocki" <rjw@sisk.pl>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
old sequence:
size ==> >4G ==> point to RAM
changed to:
>4G ==> point to RAM ==> size
some bios even leave aper to unclear, so check size at last.
To avoid reporting:
Node 0: Aperture @ 4a42000000 size 32 MB
Aperture too small (32 MB)
with this change we will get:
Node 0: Aperture @ 4a42000000 size 32 MB
Aperture beyond 4G. Ignoring.
Signed-off-by: Yinghai Lu <yinghai.lu@sun.com>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
currently when gart iommu is enabled by BIOS or previous we got
"
Checking aperture...
CPU 0: aperture @4000000 size 64MB
CPU 1: aperture @4000000 size 64MB
"
we should use use Node instead.
we will get
"
Checking aperture...
Node 0: aperture @4000000 size 64MB
Node 1: aperture @4000000 size 64MB
"
Signed-off-by: Yinghai Lu <yinghai.lu@sun.com>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
For K8 system: 4G RAM with memory hole remapping enabled, or more than
4G RAM installed.
when try to use kexec second kernel, and the first doesn't include
gart_shutdown. the second kernel could have different aper position than
the first kernel. and second kernel could use that hole as RAM that is
still used by GART set by the first kernel. esp. when try to kexec
2.6.24 with sparse mem enable from previous kernel (from RHEL 5 or SLES
10). the new kernel will use aper by GART (set by first kernel) for
vmemmap. and after new kernel setting one new GART. the position will be
real RAM. the _mapcount set is lost.
Bad page state in process 'swapper'
page:ffffe2000e600020 flags:0x0000000000000000 mapping:0000000000000000 mapcount:1 count:0
Trying to fix it up, but a reboot is needed
Backtrace:
Pid: 0, comm: swapper Not tainted 2.6.24-rc7-smp-gcdf71a10-dirty #13
Call Trace:
[<ffffffff8026401f>] bad_page+0x63/0x8d
[<ffffffff80264169>] __free_pages_ok+0x7c/0x2a5
[<ffffffff80ba75d1>] free_all_bootmem_core+0xd0/0x198
[<ffffffff80ba3a42>] numa_free_all_bootmem+0x3b/0x76
[<ffffffff80ba3461>] mem_init+0x3b/0x152
[<ffffffff80b959d3>] start_kernel+0x236/0x2c2
[<ffffffff80b9511a>] _sinittext+0x11a/0x121
and
[ffffe2000e600000-ffffe2000e7fffff] PMD ->ffff81001c200000 on node 0
phys addr is : 0x1c200000
RHEL 5.1 kernel -53 said:
PCI-DMA: aperture base @ 1c000000 size 65536 KB
new kernel said:
Mapping aperture over 65536 KB of RAM @ 3c000000
So could try to disable that GART if possible.
According to Ingo
> hm, i'm wondering, instead of modifying the GART, why dont we simply
> _detect_ whatever GART settings we have inherited, and propagate that
> into our e820 maps? I.e. if there's inconsistency, then punch that out
> from the memory maps and just dont use that memory.
>
> that way it would not matter whether the GART settings came from a [old
> or crashing] Linux kernel that has not called gart_iommu_shutdown(), or
> whether it's a BIOS that has set up an aperture hole inconsistent with
> the memory map it passed. (or the memory map we _think_ i tried to pass
> us)
>
> it would also be more robust to only read and do a memory map quirk
> based on that, than actively trying to change the GART so early in the
> bootup. Later on we have to re-enable the GART _anyway_ and have to
> punch a hole for it.
>
> and as a bonus, we would have shored up our defenses against crappy
> BIOSes as well.
add e820 modification for gart inconsistent setting.
gart_fix_e820=off could be used to disable e820 fix.
Signed-off-by: Yinghai Lu <yinghai.lu@sun.com>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
This patch renames the 4 symbols iommu_hole_init(), iommu_aperture,
iommu_aperture_allowed, iommu_aperture_disabled. All these symbols are only
used for the GART implementation of IOMMUs.
It adds and additional gart_ prefix to them.
Signed-off-by: Joerg Roedel <joerg.roedel@amd.com>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
This patch renames the include file asm-x86/iommu.h to asm-x86/gart.h to make
clear to which IOMMU implementation it belongs. The patch also adds "GART" to
the Kconfig line.
Signed-off-by: Joerg Roedel <joerg.roedel@amd.com>
Acked-by: Muli Ben-Yehuda <muli@il.ibm.com>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
