qemu-devel.nongnu.org archive mirror
 help / color / mirror / Atom feed
From: Marcel Apfelbaum <marcel@redhat.com>
To: "Michael S. Tsirkin" <mst@redhat.com>
Cc: kraxel@redhat.com, quintela@redhat.com, seabios@seabios.org,
	qemu-devel@nongnu.org, agraf@suse.de, alex.williamson@redhat.com,
	kevin@koconnor.net, qemu-ppc@nongnu.org, hare@suse.de,
	imammedo@redhat.com, amit.shah@redhat.com, pbonzini@redhat.com,
	leon.alrae@imgtec.com, aurelien@aurel32.net, rth@twiddle.net
Subject: Re: [Qemu-devel] [PATCH v5 for-2.3 27/28] apci: fix PXB behaviour if used with unsupported BIOS
Date: Tue, 10 Mar 2015 18:19:43 +0200	[thread overview]
Message-ID: <54FF199F.5010002@redhat.com> (raw)
In-Reply-To: <20150310164158-mutt-send-email-mst@redhat.com>

On 03/10/2015 05:44 PM, Michael S. Tsirkin wrote:
> On Tue, Mar 10, 2015 at 05:32:13PM +0200, Marcel Apfelbaum wrote:
>> PXB does not work with an unsupported BIOS, but should
>> not interfere with normal OS operation.
>>
>> Fix this by not adding PXB mem/IO chunks to _CRS
>> if they weren't configured by BIOS.
>>
>> Signed-off-by: Marcel Apfelbaum <marcel@redhat.com>
>> ---
>>   hw/i386/acpi-build.c | 71 +++++++++++++++++++++++++++++++---------------------
>>   1 file changed, 42 insertions(+), 29 deletions(-)
>>
>> diff --git a/hw/i386/acpi-build.c b/hw/i386/acpi-build.c
>> index 2bc8a80..416972c 100644
>> --- a/hw/i386/acpi-build.c
>> +++ b/hw/i386/acpi-build.c
>> @@ -780,6 +780,10 @@ static Aml *build_crs(PcPciInfo *pci, PciInfo *bus_info,
>>               range_base = region->value->address;
>>               range_limit = region->value->address + region->value->size - 1;
>>
>> +            if (range_base == PCI_BAR_UNMAPPED) {
>> +                continue;
>> +            }
>> +
>>               if (!strcmp(region->value->type, "io")) {
>>                   aml_append(crs,
>>                       aml_word_io(aml_min_fixed, aml_max_fixed,
>> @@ -813,41 +817,50 @@ static Aml *build_crs(PcPciInfo *pci, PciInfo *bus_info,
>>
>>               range_base = bridge_info->bus.io_range->base;
>>               range_limit = bridge_info->bus.io_range->limit;
>> -            aml_append(crs,
>> -                aml_word_io(aml_min_fixed, aml_max_fixed,
>> -                            aml_pos_decode, aml_entire_range,
>> -                            0,
>> -                            range_base,
>> -                            range_limit,
>> -                            0,
>> -                            range_limit - range_base + 1));
>> -            crs_range_insert(io_ranges, range_base, range_limit);
>> +            /* PCI Bridge I/O limit is aligned to 4K */
>
> Can't say this comment explains anything.
> So it's aligned - thus if (range_limit >> 12) is same as
> if (range_limit).
I'll use range_base == 0 as invalid value.

Thanks,
Marcel

>
> Also, generally you should test limit>base for bridges.

>
>> +            if (range_limit >> 12) {
>> +                aml_append(crs,
>> +                    aml_word_io(aml_min_fixed, aml_max_fixed,
>> +                                aml_pos_decode, aml_entire_range,
>> +                                0,
>> +                                range_base,
>> +                                range_limit,
>> +                                0,
>> +                                range_limit - range_base + 1));
>> +                crs_range_insert(io_ranges, range_base, range_limit);
>> +            }
>>
>>               range_base = bridge_info->bus.memory_range->base;
>>               range_limit = bridge_info->bus.memory_range->limit;
>> -             aml_append(crs,
>> -                aml_dword_memory(aml_pos_decode, aml_min_fixed,
>> -                                 aml_max_fixed, aml_non_cacheable,
>> -                                 aml_ReadWrite,
>> -                                 0,
>> -                                 range_base,
>> -                                 range_limit,
>> -                                 0,
>> -                                 range_limit - range_base + 1));
>> -            crs_range_insert(mem_ranges, range_base, range_limit);
>> +            /* PCI Bridge MEM limit is aligned to 1M */
>
> same comment here.
>
>> +            if (range_limit >> 20) {
>> +                aml_append(crs,
>> +                    aml_dword_memory(aml_pos_decode, aml_min_fixed,
>> +                                     aml_max_fixed, aml_non_cacheable,
>> +                                     aml_ReadWrite,
>> +                                     0,
>> +                                     range_base,
>> +                                     range_limit,
>> +                                     0,
>> +                                     range_limit - range_base + 1));
>> +                    crs_range_insert(mem_ranges, range_base, range_limit);
>> +            }
>>
>>               range_base = bridge_info->bus.prefetchable_range->base;
>>               range_limit = bridge_info->bus.prefetchable_range->limit;
>> -            aml_append(crs,
>> -                aml_dword_memory(aml_pos_decode, aml_min_fixed,
>> -                                 aml_max_fixed, aml_non_cacheable,
>> -                                 aml_ReadWrite,
>> -                                 0,
>> -                                 range_base,
>> -                                 range_limit,
>> -                                 0,
>> -                                 range_limit - range_base + 1));
>> -            crs_range_insert(mem_ranges, range_base, range_limit);
>> +            /* PCI Bridge Prefetch MEM limit is aligned to 1M */
>> +            if (range_limit >> 20) {
>> +                aml_append(crs,
>> +                    aml_dword_memory(aml_pos_decode, aml_min_fixed,
>> +                                     aml_max_fixed, aml_non_cacheable,
>> +                                     aml_ReadWrite,
>> +                                     0,
>> +                                     range_base,
>> +                                     range_limit,
>> +                                     0,
>> +                                     range_limit - range_base + 1));
>> +                crs_range_insert(mem_ranges, range_base, range_limit);
>> +            }
>>           }
>>       }
>>
>> --
>> 2.1.0

  reply	other threads:[~2015-03-10 16:20 UTC|newest]

Thread overview: 58+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2015-03-10 15:31 [Qemu-devel] [PATCH v5 for-2.3 00/28] hw/pc: implement multiple primary busses for pc machines Marcel Apfelbaum
2015-03-10 15:31 ` [Qemu-devel] [PATCH v5 for-2.3 01/28] acpi: fix aml_equal term implementation Marcel Apfelbaum
2015-03-10 15:31 ` [Qemu-devel] [PATCH v5 for-2.3 02/28] acpi: add aml_or() term Marcel Apfelbaum
2015-03-11  1:17   ` Shannon Zhao
2015-03-11 13:07     ` Marcel Apfelbaum
2015-03-10 15:31 ` [Qemu-devel] [PATCH v5 for-2.3 03/28] acpi: add aml_add() term Marcel Apfelbaum
2015-03-10 15:31 ` [Qemu-devel] [PATCH v5 for-2.3 04/28] acpi: add aml_lless() term Marcel Apfelbaum
2015-03-10 15:31 ` [Qemu-devel] [PATCH v5 for-2.3 05/28] acpi: add aml_index() term Marcel Apfelbaum
2015-03-10 15:31 ` [Qemu-devel] [PATCH v5 for-2.3 06/28] acpi: add aml_shiftleft() term Marcel Apfelbaum
2015-03-10 15:31 ` [Qemu-devel] [PATCH v5 for-2.3 07/28] acpi: add aml_shiftright() term Marcel Apfelbaum
2015-03-10 15:31 ` [Qemu-devel] [PATCH v5 for-2.3 08/28] acpi: add aml_increment() term Marcel Apfelbaum
2015-03-10 15:31 ` [Qemu-devel] [PATCH v5 for-2.3 09/28] acpi: add aml_while() term Marcel Apfelbaum
2015-03-10 15:31 ` [Qemu-devel] [PATCH v5 for-2.3 10/28] hw/acpi: add support for multiple root busses Marcel Apfelbaum
2015-03-10 15:31 ` [Qemu-devel] [PATCH v5 for-2.3 11/28] hw/apci: add _PRT method for extra PCI " Marcel Apfelbaum
2015-03-10 16:41   ` Michael S. Tsirkin
2015-03-10 15:31 ` [Qemu-devel] [PATCH v5 for-2.3 12/28] hw/acpi: add _CRS method for extra " Marcel Apfelbaum
2015-03-10 15:38   ` Michael S. Tsirkin
2015-03-10 16:17     ` Marcel Apfelbaum
2015-03-10 15:31 ` [Qemu-devel] [PATCH v5 for-2.3 13/28] hw/acpi: remove from root bus 0 the crs resources used by other busses Marcel Apfelbaum
2015-03-10 15:32 ` [Qemu-devel] [PATCH v5 for-2.3 14/28] hw/pci: move pci bus related code to separate files Marcel Apfelbaum
2015-03-10 15:32 ` [Qemu-devel] [PATCH v5 for-2.3 15/28] hw/pci: made pci_bus_is_root a PCIBusClass method Marcel Apfelbaum
2015-03-10 15:32 ` [Qemu-devel] [PATCH v5 for-2.3 16/28] hw/pci: made pci_bus_num " Marcel Apfelbaum
2015-03-10 15:32 ` [Qemu-devel] [PATCH v5 for-2.3 17/28] hw/pci: introduce TYPE_PCI_MAIN_HOST_BRIDGE interface Marcel Apfelbaum
2015-03-10 15:32 ` [Qemu-devel] [PATCH v5 for-2.3 18/28] hw/pci: removed 'rootbus nr is 0' assumption from qmp_pci_query Marcel Apfelbaum
2015-03-10 15:32 ` [Qemu-devel] [PATCH v5 for-2.3 19/28] hw/pci: implement iteration over multiple host bridges Marcel Apfelbaum
2015-03-10 16:39   ` Michael S. Tsirkin
2015-03-10 15:32 ` [Qemu-devel] [PATCH v5 for-2.3 20/28] hw/pci: introduce PCI Expander Bridge (PXB) Marcel Apfelbaum
2015-03-10 15:32 ` [Qemu-devel] [PATCH v5 for-2.3 21/28] hw/pci: inform bios if the system has more than one pci bridge Marcel Apfelbaum
2015-03-10 16:36   ` Michael S. Tsirkin
2015-03-10 15:32 ` [Qemu-devel] [PATCH v5 for-2.3 22/28] hw/pci: piix - suport multiple host bridges Marcel Apfelbaum
2015-03-10 16:22   ` Michael S. Tsirkin
2015-03-10 15:32 ` [Qemu-devel] [PATCH v5 for-2.3 23/28] hw/pxb: add map_irq func Marcel Apfelbaum
2015-03-10 16:43   ` Michael S. Tsirkin
2015-03-16 12:11     ` Marcel Apfelbaum
2015-03-16 15:27       ` Michael S. Tsirkin
2015-03-10 15:32 ` [Qemu-devel] [PATCH v5 for-2.3 24/28] hw/pci_bus: add support for NUMA nodes Marcel Apfelbaum
2015-03-10 15:32 ` [Qemu-devel] [PATCH v5 for-2.3 25/28] hw/pxb: add numa_node parameter Marcel Apfelbaum
2015-03-10 15:32 ` [Qemu-devel] [PATCH v5 for-2.3 26/28] acpi: restrict the aml emission to PXB host bridges Marcel Apfelbaum
2015-03-10 15:41   ` Michael S. Tsirkin
2015-03-10 16:18     ` Marcel Apfelbaum
2015-03-10 15:32 ` [Qemu-devel] [PATCH v5 for-2.3 27/28] apci: fix PXB behaviour if used with unsupported BIOS Marcel Apfelbaum
2015-03-10 15:44   ` Michael S. Tsirkin
2015-03-10 16:19     ` Marcel Apfelbaum [this message]
2015-03-10 16:21       ` Michael S. Tsirkin
2015-03-10 15:32 ` [Qemu-devel] [PATCH v5 for-2.3 28/28] docs: Add PXB documentation Marcel Apfelbaum
2015-03-10 15:47   ` Michael S. Tsirkin
2015-03-10 16:21     ` Marcel Apfelbaum
2015-03-10 17:42       ` Michael S. Tsirkin
2015-03-16 12:16         ` Marcel Apfelbaum
2015-03-16 15:28           ` Michael S. Tsirkin
2015-03-16 15:47             ` Marcel Apfelbaum
2015-03-11 13:32 ` [Qemu-devel] [PATCH v5 for-2.3 00/28] hw/pc: implement multiple primary busses for pc machines Gerd Hoffmann
2015-03-11 13:44   ` Marcel Apfelbaum
2015-03-11 13:51     ` Gerd Hoffmann
2015-03-11 14:01       ` Marcel Apfelbaum
2015-03-11 14:02         ` Michael S. Tsirkin
2015-03-11 14:12 ` Gerd Hoffmann
2015-03-11 14:14   ` Marcel Apfelbaum

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=54FF199F.5010002@redhat.com \
    --to=marcel@redhat.com \
    --cc=agraf@suse.de \
    --cc=alex.williamson@redhat.com \
    --cc=amit.shah@redhat.com \
    --cc=aurelien@aurel32.net \
    --cc=hare@suse.de \
    --cc=imammedo@redhat.com \
    --cc=kevin@koconnor.net \
    --cc=kraxel@redhat.com \
    --cc=leon.alrae@imgtec.com \
    --cc=mst@redhat.com \
    --cc=pbonzini@redhat.com \
    --cc=qemu-devel@nongnu.org \
    --cc=qemu-ppc@nongnu.org \
    --cc=quintela@redhat.com \
    --cc=rth@twiddle.net \
    --cc=seabios@seabios.org \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).