[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: [Xen-devel] [PATCH] xen/balloon: Mark unallocated host memory as UNUSABLE



On 12/12/17 23:51, Boris Ostrovsky wrote:
> Commit f5775e0b6116 ("x86/xen: discard RAM regions above the maximum
> reservation") left host memory not assigned to dom0 as available for
> memory hotplug.
> 
> Unfortunately this also meant that those regions could be used by
> others. Specifically, commit fa564ad96366 ("x86/PCI: Enable a 64bit BAR
> on AMD Family 15h (Models 00-1f, 30-3f, 60-7f)") may try to map those
> addresses as MMIO.
> 
> To prevent this mark unallocated host memory as E820_TYPE_UNUSABLE (thus
> effectively reverting f5775e0b6116) and keep track of that region as
> a hostmem resource that can be used for the hotplug.
> 
> Signed-off-by: Boris Ostrovsky <boris.ostrovsky@xxxxxxxxxx>
> ---
> I don't see /proc/meminfo reporting the hotplugged memory (although
> internal data such as max_pfn is updated properly). Need to look at
> hotplug code some more. But then I didn't see meminfo changing with
> existing code either.
> 
> 
>  arch/x86/xen/enlighten.c | 69 
> ++++++++++++++++++++++++++++++++++++++++++++++++
>  arch/x86/xen/setup.c     |  6 ++---
>  drivers/xen/balloon.c    | 65 ++++++++++++++++++++++++++++++++++++++-------
>  3 files changed, 127 insertions(+), 13 deletions(-)
> 
> diff --git a/arch/x86/xen/enlighten.c b/arch/x86/xen/enlighten.c
> index d669e9d..19223b9 100644
> --- a/arch/x86/xen/enlighten.c
> +++ b/arch/x86/xen/enlighten.c
> @@ -3,6 +3,7 @@
>  
>  #include <xen/features.h>
>  #include <xen/page.h>
> +#include <xen/interface/memory.h>
>  
>  #include <asm/xen/hypercall.h>
>  #include <asm/xen/hypervisor.h>
> @@ -331,3 +332,71 @@ void xen_arch_unregister_cpu(int num)
>  }
>  EXPORT_SYMBOL(xen_arch_unregister_cpu);
>  #endif
> +
> +#ifdef CONFIG_XEN_BALLOON_MEMORY_HOTPLUG
> +void __init arch_xen_balloon_init(struct resource *hostmem_resource)
> +{
> +     struct xen_memory_map memmap;
> +     int rc, i, last_guest_ram;

i and last_guest_ram should be unsigned int

> +     unsigned long max_addr = max_pfn << PAGE_SHIFT;

This might overflow for 32 bit builds

> +     struct e820_table *xen_e820_table;
> +     struct e820_entry *entry;
> +     struct resource *res = NULL;
> +
> +     if (!xen_initial_domain())
> +             return;
> +
> +     xen_e820_table = kzalloc(sizeof(*xen_e820_table), GFP_KERNEL);
> +     if (!xen_e820_table) {
> +             pr_warn("%s: Out of memory\n", __func__);
> +             return;
> +     }
> +
> +     memmap.nr_entries = ARRAY_SIZE(xen_e820_table->entries);
> +     set_xen_guest_handle(memmap.buffer, xen_e820_table->entries);
> +     rc = HYPERVISOR_memory_op(XENMEM_machine_memory_map, &memmap);
> +     if (rc) {
> +             pr_warn("%s: Can't read host e820 (%d)\n", __func__, rc);
> +             goto out;
> +     }
> +
> +     last_guest_ram = i = 0;
> +     while (xen_e820_table->entries[i].addr < max_addr) {
> +             if (xen_e820_table->entries[i].type == E820_TYPE_RAM)
> +                     last_guest_ram = i;
> +             i++;

Check for i < memmap.nr_entries?

> +     }
> +
> +     entry = &xen_e820_table->entries[last_guest_ram];
> +     if (max_addr >= entry->addr + entry->size)
> +             goto out; /* No unallocated host RAM. */
> +
> +     hostmem_resource->start = max_addr;
> +     hostmem_resource->end = entry->addr + entry->size;
> +     for (; i < memmap.nr_entries; i++) {
> +             entry = &xen_e820_table->entries[i];
> +             if (entry->type == E820_TYPE_RAM)

Shouldn't that be != ?

> +                     continue;
> +
> +             res = kzalloc(sizeof(*res), GFP_KERNEL);
> +             if (!res) {
> +                     pr_warn("%s: Out of memory\n", __func__);

Don't print error message in case of allocation failures.


Juergen

_______________________________________________
Xen-devel mailing list
Xen-devel@xxxxxxxxxxxxxxxxxxxx
https://lists.xenproject.org/mailman/listinfo/xen-devel

 


Rackspace

Lists.xenproject.org is hosted with RackSpace, monitoring our
servers 24x7x365 and backed by RackSpace's Fanatical Support®.