Re: [PATCH v2] fuzz: map all BARs and enable PCI devices

[Thomas Huth]

On 15/12/2020 20.08, Alexander Bulekov wrote:
> Prior to this patch, the fuzzer found inputs to map PCI device BARs and
> enable the device. While it is nice that the fuzzer can do this, it
> added significant overhead, since the fuzzer needs to map all the
> BARs (regenerating the memory topology), at the start of each input.
> With this patch, we do this once, before fuzzing, mitigating some of
> this overhead.
> 
> Signed-off-by: Alexander Bulekov <alxndr@bu.edu>
> Reviewed-by: Darren Kenny <darren.kenny@oracle.com>
> ---
> 
> v2: do not mix statements and declarations
> 
>  tests/qtest/fuzz/generic_fuzz.c | 23 +++++++++++++++++++++++
>  1 file changed, 23 insertions(+)
> 
> diff --git a/tests/qtest/fuzz/generic_fuzz.c b/tests/qtest/fuzz/generic_fuzz.c
> index 07ad690683..92b598dea8 100644
> --- a/tests/qtest/fuzz/generic_fuzz.c
> +++ b/tests/qtest/fuzz/generic_fuzz.c
> @@ -16,6 +16,7 @@
>  
>  #include "hw/core/cpu.h"
>  #include "tests/qtest/libqos/libqtest.h"
> +#include "tests/qtest/libqos/pci-pc.h"
>  #include "fuzz.h"
>  #include "fork_fuzz.h"
>  #include "exec/address-spaces.h"
> @@ -762,10 +763,28 @@ static int locate_fuzz_objects(Object *child, void 
> *opaque)
>      return 0;
>  }
>  
> +
> +static void pci_enum(gpointer pcidev, gpointer bus)
> +{
> +    PCIDevice *dev = pcidev;
> +    QPCIDevice *qdev;
> +    int i;
> +
> +    qdev = qpci_device_find(bus, dev->devfn);
> +    g_assert(qdev != NULL);
> +    for (i = 0; i < 6; i++) {
> +        if (dev->io_regions[i].size) {
> +            qpci_iomap(qdev, i, NULL);
> +        }
> +    }
> +    qpci_device_enable(qdev);
> +}

This new code just triggered a leak error in the gitlab-CI for me:

 https://gitlab.com/huth/qemu/-/jobs/920543745#L309

Could you please have a look?

 Thanks,
  Thomas