On Fri, 2010-07-30 at 20:42 +0000, H. Peter Anvin wrote:
> On 07/30/2010 09:53 AM, David Woodhouse wrote:
> > On Thu, 2010-07-29 at 21:41 -0700, Eric W. Biederman wrote:
> >> There isn't a bus master shut off at the core level.
> >
> > Effectively, there is if you have an IOMMU.
>
> With the "core level" I meant Linux kernel code, as opposed to hardware
> level which is slightly different; I meant it would make sense to at
> least set the bus master control bit (PCI_COMMAND_MASTER) to zero before
> kexec.
Before kexec patch for ia64 was merged into mainline kernel, Zou Nan Hai
and I had added a device_shootdown() routine to arch/ia64/kernel/crash.c
that was called from machine_crash_shutdown(). device_shootdown() did
exactly what you are proposing:
+static void device_shootdown(void)
+{
+ struct pci_dev *dev;
+ irq_desc_t *desc;
+ u16 pci_command;
+
+ list_for_each_entry(dev, &pci_devices, global_list) {
+ desc = irq_descp(dev->irq);
+ if (!desc->action)
+ continue;
+ pci_read_config_word(dev, PCI_COMMAND, &pci_command);
+ if (pci_command & PCI_COMMAND_MASTER) {
+ pci_command &= ~PCI_COMMAND_MASTER;
+ pci_write_config_word(dev, PCI_COMMAND, pci_command);
+ }
+ disable_irq_nosync(dev->irq);
+ desc->handler->end(dev->irq);
+ }
+}
There were some discussions regarding this and this code was removed by
the time it was merged into mainline kernel. I can't remember the
details of why. I remember one report of kernel hang on kexec that
seemed to happen in device_shootdown(). I will look for any discussion
threads I can find.
--
Khalid
====================================================================
Khalid Aziz Telco Platform Software, ISB
(970)898-9214 Hewlett-Packard
khalid.aziz at hp.com Fort Collins, CO
"The Linux kernel is subject to relentless development"
- Alessandro Rubini
