From: ebiederm@xmission.com (Eric W. Biederman)
To: Neil Horman <nhorman@tuxdriver.com>
Cc: Neil Horman <nhorman@redhat.com>,
Yinghai Lu <yhlu.kernel@gmail.com>,
Ben Woodard <woodard@redhat.com>,
kexec@lists.infradead.org, linux-kernel@vger.kernel.org,
Andi Kleen <ak@suse.de>,
hbabu@us.ibm.com, Andi Kleen <andi@firstfloor.org>
Subject: Re: [PATCH] kexec: force x86_64 arches to boot kdump kernels on boot cpu
Date: Mon, 10 Dec 2007 21:48:11 -0700 [thread overview]
Message-ID: <m1ejdt25x0.fsf@ebiederm.dsl.xmission.com> (raw)
In-Reply-To: <20071211034349.GA3635@localhost.localdomain> (Neil Horman's message of "Mon, 10 Dec 2007 22:43:49 -0500")
Neil Horman <nhorman@tuxdriver.com> writes:
Almost there.
> On Mon, Dec 10, 2007 at 06:08:03PM -0700, Eric W. Biederman wrote:
>> Neil Horman <nhorman@tuxdriver.com> writes:
>>
> <snip>
>>
>> Ok. This test is broken. Please remove the == 1. You are looking
>> for == (1 << 18). So just saying: "if (htcfg & (1 << 18))" should be clearer.
>>
> Fixed. Thanks!
>
>> > + printk(KERN_INFO "Detected use of extended apic ids on hypertransport
> bus\n");
>> > + if ((htcfg & (1 << 17)) == 0) {
>> > + printk(KERN_INFO "Enabling hypertransport extended apic interrupt
>> > broadcast\n");
>> > + htcfg |= (1 << 17);
>> > + write_pci_config(num, slot, func, 0x68, htcfg);
>> > + }
>> > + }
>> > +
>> > +}
>>
>> The rest of this quirk looks fine, include the fact it is only intended
>> to be applied to PCI_VENDOR_ID_AMD PCI_DEVICE_ID_AMD_K8_NB.
>>
> Copy that.
>
>>
>> For what is below I don't like the way the infrastructure has been
>> extended as what you are doing quickly devolves into a big mess.
>>
>> Please extend struct chipset to be something like:
>> struct chipset {
>> u16 vendor;
>> u16 device;
>> u32 class, class_mask;
>> void (*f)(void);
>> };
>>
>> And then the test for matching the chipset can be something like:
>> if ((id->vendor == PCI_ANY_ID || id->vendor == dev->vendor) &&
>> (id->device == PCI_ANY_ID || id->device == dev->device) &&
>> !((id->class ^ dev->class) & id->class_mask))
>>
>> Essentially a subset of pci_match_one_device from drivers/pci/pci.h
>>
>> That way you don't need to increase the number of tables or the
>> number of passes through the pci busses, just update the early_qrk
>> table with a few more bits of information.
>>
> copy that. Fixed. Thanks!
>
>> The extended form should be much more maintainable in the long
>> run. Given that we may want this before we enable the timer
>> which is very early doing this in the pci early quirks seems
>> to make sense.
>>
>> Eric
>
>
> New patch attached, with suggestions incorporated.
>
> Thanks & regards
> Neil
>
> Signed-off-by: Neil Horman <nhorman@tuxdriver.com>
>
>
> early-quirks.c | 82 ++++++++++++++++++++++++++++++++++++++++++++++++++-------
> 1 file changed, 73 insertions(+), 9 deletions(-)
>
>
>
> diff --git a/arch/x86/kernel/early-quirks.c b/arch/x86/kernel/early-quirks.c
> index 88bb83e..4b0cee1 100644
> --- a/arch/x86/kernel/early-quirks.c
> +++ b/arch/x86/kernel/early-quirks.c
> @@ -44,6 +44,50 @@ static int __init nvidia_hpet_check(struct acpi_table_header
> *header)
> #endif /* CONFIG_X86_IO_APIC */
> #endif /* CONFIG_ACPI */
>
> +static void __init fix_hypertransport_config(int num, int slot, int func)
> +{
> + u32 htcfg;
> + /*
> + *we found a hypertransport bus
> + *make sure that are broadcasting
> + *interrupts to all cpus on the ht bus
> + *if we're using extended apic ids
> + */
> + htcfg = read_pci_config(num, slot, func, 0x68);
> + if (htcfg & (1 << 18)) {
> + printk(KERN_INFO "Detected use of extended apic ids on hypertransport bus\n");
> + if ((htcfg & (1 << 17)) == 0) {
> + printk(KERN_INFO "Enabling hypertransport extended apic interrupt
> broadcast\n");
> + htcfg |= (1 << 17);
> + write_pci_config(num, slot, func, 0x68, htcfg);
> + }
> + }
> +
> +}
> +
> +static void __init check_hypertransport_config()
> +{
> + int num, slot, func;
> + u32 device, vendor;
> + func = 0;
> + for (num = 0; num < 32; num++) {
> + for (slot = 0; slot < 32; slot++) {
> + vendor = read_pci_config(num,slot,func,
> + PCI_VENDOR_ID);
> + device = read_pci_config(num,slot,func,
> + PCI_DEVICE_ID);
> + vendor &= 0x0000ffff;
> + device >>= 16;
> + if ((vendor == PCI_VENDOR_ID_AMD) &&
> + (device == PCI_DEVICE_ID_AMD_K8_NB))
> + fix_hypertransport_config(num,slot,func);
> + }
> + }
> +
> + return;
> +
> +}
We should not need check_hypertransport_config as the generic loop
now does the work for us.
> +
> static void __init nvidia_bugs(void)
> {
> #ifdef CONFIG_ACPI
> @@ -83,15 +127,25 @@ static void __init ati_bugs(void)
> #endif
> }
>
> +static void __init amd_host_bugs(void)
> +{
> + printk(KERN_CRIT "IN AMD_HOST_BUGS\n");
> + check_hypertransport_config();
> +}
Likewise this function is unneeded and the printk is likely confusing
for users.
> struct chipset {
> u16 vendor;
> + u16 device;
> + u32 class;
> + u32 class_mask;
> void (*f)(void);
> };
>
> static struct chipset early_qrk[] __initdata = {
> - { PCI_VENDOR_ID_NVIDIA, nvidia_bugs },
> - { PCI_VENDOR_ID_VIA, via_bugs },
> - { PCI_VENDOR_ID_ATI, ati_bugs },
> + { PCI_VENDOR_ID_NVIDIA, PCI_ANY_ID, PCI_CLASS_BRIDGE_PCI, PCI_ANY_ID,
> nvidia_bugs },
> + { PCI_VENDOR_ID_VIA, PCI_ANY_ID, PCI_CLASS_BRIDGE_PCI, PCI_ANY_ID, via_bugs },
> + { PCI_VENDOR_ID_ATI, PCI_ANY_ID, PCI_CLASS_BRIDGE_PCI, PCI_ANY_ID, ati_bugs },
> + { PCI_VENDOR_ID_AMD, PCI_DEVICE_ID_AMD_K8_NB, PCI_CLASS_BRIDGE_HOST,
> PCI_ANY_ID, amd_host_bugs },
> {}
So make that fix_hypertransport_config and we should be good.
> };
>
> @@ -108,25 +162,35 @@ void __init early_quirks(void)
> for (func = 0; func < 8; func++) {
> u32 class;
> u32 vendor;
> + u32 device;
> u8 type;
> int i;
> +
> class = read_pci_config(num,slot,func,
> PCI_CLASS_REVISION);
> if (class == 0xffffffff)
> break;
>
> - if ((class >> 16) != PCI_CLASS_BRIDGE_PCI)
> - continue;
> + class >>= 16;
>
> vendor = read_pci_config(num, slot, func,
> PCI_VENDOR_ID);
> vendor &= 0xffff;
>
> - for (i = 0; early_qrk[i].f; i++)
> - if (early_qrk[i].vendor == vendor) {
> - early_qrk[i].f();
> - return;
> + device = read_pci_config(num, slot, func,
> + PCI_DEVICE_ID);
> + device >>= 16;
We don't need to shift device. Although we can do:
device_vendor = read_pci_config(num, slot, func, PCI_VENDOR_ID);
device = device_vendor >> 16;
vendor = device_vendor & 0xffff;
Assuming the early read_pci_config is limited to 32bit reads.
> + for(i=0;early_qrk[i].f != NULL;i++) {
> + if (((early_qrk[i].vendor == PCI_ANY_ID) ||
> + (early_qrk[i].vendor == vendor)) &&
> + ((early_qrk[i].device == PCI_ANY_ID) ||
> + (early_qrk[i].device == device)) &&
> + ((early_qrk[i].class ^ class) &
> + early_qrk[i].class_mask)) {
> + early_qrk[i].f();
> }
> + }
>
> type = read_pci_config_byte(num, slot, func,
> PCI_HEADER_TYPE);
next prev parent reply other threads:[~2007-12-11 4:51 UTC|newest]
Thread overview: 100+ messages / expand[flat|nested] mbox.gz Atom feed top
2007-11-27 1:47 Neil Horman
2007-11-27 4:12 ` Eric W. Biederman
2007-11-27 13:13 ` Neil Horman
2007-11-27 13:28 ` Eric W. Biederman
2007-11-27 13:45 ` Andi Kleen
2007-11-27 14:28 ` Neil Horman
2007-11-27 14:43 ` Andi Kleen
2007-11-27 14:48 ` Neil Horman
2007-11-27 15:24 ` Andi Kleen
2007-11-27 15:30 ` Eric W. Biederman
2007-11-27 16:45 ` Neil Horman
2007-11-27 20:50 ` Ben Woodard
2007-11-27 21:05 ` Neil Horman
2007-11-27 22:38 ` Eric W. Biederman
2007-11-27 23:15 ` Ben Woodard
2007-11-28 0:15 ` Eric W. Biederman
2007-11-27 23:40 ` Neil Horman
2007-11-28 0:43 ` Eric W. Biederman
2007-11-28 15:54 ` Neil Horman
2007-11-27 14:56 ` Eric W. Biederman
2007-11-27 15:34 ` Neil Horman
2007-11-27 18:41 ` Ben Woodard
2007-11-27 19:42 ` Neil Horman
2007-11-27 20:00 ` Vivek Goyal
2007-11-27 20:52 ` Neil Horman
2007-11-27 22:24 ` Andi Kleen
2007-11-27 23:24 ` Ben Woodard
2007-11-27 23:56 ` Andi Kleen
2007-11-28 15:36 ` Vivek Goyal
2007-11-28 16:02 ` Neil Horman
2007-11-28 17:36 ` Eric W. Biederman
2007-11-28 18:16 ` Neil Horman
2007-11-28 19:05 ` Vivek Goyal
2007-11-28 19:42 ` Eric W. Biederman
2007-11-28 21:09 ` Neil Horman
2007-11-28 23:27 ` Eric W. Biederman
2007-11-30 2:16 ` Ben Woodard
2007-11-30 2:54 ` Eric W. Biederman
2007-11-30 8:59 ` Yinghai Lu
2007-11-30 14:35 ` Vivek Goyal
2007-11-30 14:32 ` Neil Horman
2007-11-30 2:12 ` Ben Woodard
2007-11-30 14:42 ` Vivek Goyal
2007-11-30 14:51 ` Neil Horman
2007-12-06 21:39 ` Neil Horman
2007-12-06 22:11 ` Vivek Goyal
2007-12-07 0:10 ` Neil Horman
2007-12-07 14:39 ` Vivek Goyal
2007-12-07 14:53 ` Neil Horman
2007-12-07 15:16 ` Vivek Goyal
2007-12-07 15:53 ` Neil Horman
2007-12-07 18:46 ` Eric W. Biederman
2007-12-07 0:33 ` Eric W. Biederman
2007-12-07 2:04 ` Neil Horman
2007-12-07 8:50 ` Yinghai Lu
2007-12-07 9:22 ` Yinghai Lu
2007-12-07 14:21 ` Neil Horman
2007-12-07 17:58 ` Neil Horman
2007-12-07 19:19 ` yhlu
2007-12-07 20:13 ` Neil Horman
2007-12-10 15:39 ` Neil Horman
2007-12-10 16:20 ` Vivek Goyal
2007-12-11 1:17 ` Eric W. Biederman
2007-12-11 1:08 ` Eric W. Biederman
2007-12-11 3:43 ` Neil Horman
2007-12-11 4:48 ` Eric W. Biederman [this message]
2007-12-11 6:31 ` Yinghai Lu
2007-12-11 14:39 ` Neil Horman
2007-12-11 15:29 ` Eric W. Biederman
2007-12-11 18:00 ` Yinghai Lu
2007-12-11 18:29 ` Neil Horman
2007-12-11 18:45 ` Yinghai Lu
2007-12-11 18:22 ` Neil Horman
2007-12-11 18:46 ` Eric W. Biederman
2007-12-11 19:24 ` Neil Horman
2007-12-11 19:51 ` Yinghai Lu
2007-12-11 20:59 ` Neil Horman
2007-12-12 0:16 ` Ben Woodard
2007-12-12 0:52 ` Neil Horman
2007-12-12 1:07 ` Yinghai Lu
2007-12-12 8:43 ` [PATCH] k8: Enable legacy irqs with extended cpu ids Eric W. Biederman
2007-12-12 14:21 ` [PATCH] kexec: force x86_64 arches to boot kdump kernels on boot cpu Andi Kleen
2007-12-12 15:55 ` Neil Horman
2007-12-12 16:07 ` Andi Kleen
2007-12-12 19:43 ` Eric W. Biederman
2007-12-12 20:22 ` Neil Horman
2007-12-12 21:32 ` Eric W. Biederman
2007-12-13 14:39 ` Neil Horman
2007-12-13 15:16 ` Andi Kleen
2007-12-13 15:32 ` Neil Horman
2007-12-17 11:38 ` Neil Horman
2007-12-18 0:13 ` Eric W. Biederman
2007-12-17 15:16 ` Ingo Molnar
2007-12-17 15:47 ` Neil Horman
2007-12-07 18:36 ` Eric W. Biederman
2007-12-07 18:48 ` Neil Horman
2007-11-27 13:53 ` Neil Horman
2007-11-27 10:55 ` Andi Kleen
2007-11-27 11:19 ` Eric W. Biederman
2007-11-27 13:28 ` Neil Horman
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=m1ejdt25x0.fsf@ebiederm.dsl.xmission.com \
--to=ebiederm@xmission.com \
--cc=ak@suse.de \
--cc=andi@firstfloor.org \
--cc=hbabu@us.ibm.com \
--cc=kexec@lists.infradead.org \
--cc=linux-kernel@vger.kernel.org \
--cc=nhorman@redhat.com \
--cc=nhorman@tuxdriver.com \
--cc=woodard@redhat.com \
--cc=yhlu.kernel@gmail.com \
/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
all inboxes | Powered by JetHome®