Re: [PATCH] spapr: Fix buffer overflow in spapr_numa_associativity_init()

[Greg Kurz]

On Fri, 18 Dec 2020 16:16:49 +0100
Philippe Mathieu-Daudé <philmd@redhat.com> wrote:

> On 12/18/20 2:53 PM, Greg Kurz wrote:
> > Running a guest with 128 NUMA nodes crashes QEMU:
> > 
> > ../../util/error.c:59: error_setv: Assertion `*errp == NULL' failed.
> > 
> > The crash happens when setting the FWNMI migration blocker:
> > 
> > 2861            if (spapr_get_cap(spapr, SPAPR_CAP_FWNMI) == SPAPR_CAP_ON) {
> > 2862                /* Create the error string for live migration blocker */
> > 2863                error_setg(&spapr->fwnmi_migration_blocker,
> > 2864                    "A machine check is being handled during migration. 
> > The handler"
> > 2865                    "may run and log hardware error on the 
> > destination");
> > 2866            }
> > 
> > Inspection reveals that papr->fwnmi_migration_blocker isn't NULL:
> > 
> > (gdb) p spapr->fwnmi_migration_blocker
> > $1 = (Error *) 0x8000000004000000
> > 
> > Since this is the only place where papr->fwnmi_migration_blocker is
> > set, this means someone wrote there in our back. Further analysis
> > points to spapr_numa_associativity_init(), especially the part
> > that initializes the associative arrays for NVLink GPUs:
> > 
> >     max_nodes_with_gpus = nb_numa_nodes + NVGPU_MAX_NUM;
> > 
> > ie. max_nodes_with_gpus = 128 + 6, but the array isn't sized to
> > accommodate the 6 extra nodes:
> > 
> > #define MAX_NODES 128
> > 
> > struct SpaprMachineState {
> >     .
> >     .
> >     .
> >     uint32_t numa_assoc_array[MAX_NODES][NUMA_ASSOC_SIZE];
> > 
> >     Error *fwnmi_migration_blocker;
> > };
> > 
> > and the following loops happily overwrite spapr->fwnmi_migration_blocker,
> > and probably more:
> > 
> >     for (i = nb_numa_nodes; i < max_nodes_with_gpus; i++) {
> 
> Along with your fix you can also remove max_nodes_with_gpus
> and use the safer ARRAY_SIZE(numa_assoc_array)...
> 

No. We want to fill the array according to the specified NUMA topology, ie.

        [ nb_numa_nodes, nb_numa_nodes + NVGPU_MAX_NUM [

not

        [ nb_numa_nodes, MAX_NODES + NVGPU_MAX_NUM [

> >         spapr->numa_assoc_array[i][0] = 
> > cpu_to_be32(MAX_DISTANCE_REF_POINTS);
> > 
> >         for (j = 1; j < MAX_DISTANCE_REF_POINTS; j++) {
> >             uint32_t gpu_assoc = smc->pre_5_1_assoc_refpoints ?
> >                                  SPAPR_GPU_NUMA_ID : cpu_to_be32(i);
> >             spapr->numa_assoc_array[i][j] = gpu_assoc;
> >         }
> > 
> >         spapr->numa_assoc_array[i][MAX_DISTANCE_REF_POINTS] = 
> > cpu_to_be32(i);
> >     }
> > 
> > Fix the size of the array. This requires "hw/ppc/spapr.h" to see
> > NVGPU_MAX_NUM. Including "hw/pci-host/spapr.h" introduces a
> > circular dependency that breaks the build, so this moves the
> > definition of NVGPU_MAX_NUM to "hw/ppc/spapr.h" instead.
> > 
> > Reported-by: Min Deng <mdeng@redhat.com>
> > BugLink: https://bugzilla.redhat.com/show_bug.cgi?id=1908693
> > Fixes: dd7e1d7ae431 ("spapr_numa: move NVLink2 associativity handling to 
> > spapr_numa.c")
> > Cc: danielhb413@gmail.com
> > Signed-off-by: Greg Kurz <groug@kaod.org>
> > ---
> >  include/hw/pci-host/spapr.h |    2 --
> >  include/hw/ppc/spapr.h      |    5 ++++-
> >  2 files changed, 4 insertions(+), 3 deletions(-)
> > 
> > diff --git a/include/hw/pci-host/spapr.h b/include/hw/pci-host/spapr.h
> > index 4f58f0223b56..bd014823a933 100644
> > --- a/include/hw/pci-host/spapr.h
> > +++ b/include/hw/pci-host/spapr.h
> > @@ -115,8 +115,6 @@ struct SpaprPhbState {
> >  #define SPAPR_PCI_NV2RAM64_WIN_BASE  SPAPR_PCI_LIMIT
> >  #define SPAPR_PCI_NV2RAM64_WIN_SIZE  (2 * TiB) /* For up to 6 GPUs 256GB 
> > each */
> >  
> > -/* Max number of these GPUsper a physical box */
> > -#define NVGPU_MAX_NUM                6
> >  /* Max number of NVLinks per GPU in any physical box */
> >  #define NVGPU_MAX_LINKS              3
> >  
> > diff --git a/include/hw/ppc/spapr.h b/include/hw/ppc/spapr.h
> > index 06a5b4259f20..1cc19575f548 100644
> > --- a/include/hw/ppc/spapr.h
> > +++ b/include/hw/ppc/spapr.h
> > @@ -112,6 +112,9 @@ typedef enum {
> >  #define NUMA_ASSOC_SIZE            (MAX_DISTANCE_REF_POINTS + 1)
> >  #define VCPU_ASSOC_SIZE            (NUMA_ASSOC_SIZE + 1)
> >  
> > +/* Max number of these GPUsper a physical box */
> > +#define NVGPU_MAX_NUM                6
> > +
> >  typedef struct SpaprCapabilities SpaprCapabilities;
> >  struct SpaprCapabilities {
> >      uint8_t caps[SPAPR_CAP_NUM];
> > @@ -240,7 +243,7 @@ struct SpaprMachineState {
> >      unsigned gpu_numa_id;
> >      SpaprTpmProxy *tpm_proxy;
> >  
> > -    uint32_t numa_assoc_array[MAX_NODES][NUMA_ASSOC_SIZE];
> > +    uint32_t numa_assoc_array[MAX_NODES + NVGPU_MAX_NUM][NUMA_ASSOC_SIZE];
> >  
> >      Error *fwnmi_migration_blocker;
> >  };
> > 
> > 
> > 
>