[Intel-gfx] [igt-dev] [PATCH v2 i-g-t] lib/intel_mmio: Fix mmapped resources not unmapped on fini
Janusz Krzysztofik
janusz.krzysztofik at linux.intel.com
Fri Mar 4 16:20:19 UTC 2022
Hi Kamil,
Thanks for review.
On Friday, 4 March 2022 16:14:05 CET Kamil Konieczny wrote:
> Hi Janusz,
>
> Dnia 2022-03-01 at 15:07:55 +0100, Janusz Krzysztofik napisał(a):
> > Commit 5f3cfa485eb4 ("lib: Use safe wrappers around libpciaccess
> > initialization functions") took care of not leaking memory allocated by
> > pci_system_init() but didn't take care of users potentially attempting to
> > reinitialize global data maintained by libpciaccess. For example,
> > intel_register_access_init() mmaps device's PCI BAR0 resource with
> > pci_device_map_range() but intel_register_access_fini() doesn't unmap it
> > and next call to intel_register_access_init() fails on attempt to mmap it
> > again with pci_device_map_range().
> ------ ^
> imho you can cut here, no need to repeat it twice.
OK
> >
> > Fix it, and also provide intel_mmio_umap_*() counterparts to public
> -------------------------------------- ^
> s/umap/unmap/
Thanks.
> > functions intel_mmio_use_pci_bar() and intel_mmio_use_dump_file().
> >
> > v2: apply last minute fixes, cached but unfortunately not committed before
> > sending
> >
> > Signed-off-by: Janusz Krzysztofik <janusz.krzysztofik at linux.intel.com>
> > ---
> > lib/intel_io.h | 4 +++
> > lib/intel_mmio.c | 67 ++++++++++++++++++++++++++++++++++++++++++------
> > 2 files changed, 63 insertions(+), 8 deletions(-)
> >
> > diff --git a/lib/intel_io.h b/lib/intel_io.h
> > index 1cfe4fb6b9..ea2649d9bc 100644
> > --- a/lib/intel_io.h
> > +++ b/lib/intel_io.h
> > @@ -49,6 +49,8 @@ struct intel_register_map {
> >
> > struct intel_mmio_data {
> > void *igt_mmio;
> > + size_t mmio_size;
> > + struct pci_device *dev;
> > struct intel_register_map map;
> > uint32_t pci_device_id;
> > int key;
> > @@ -57,7 +59,9 @@ struct intel_mmio_data {
> >
> > void intel_mmio_use_pci_bar(struct intel_mmio_data *mmio_data,
> > struct pci_device *pci_dev);
> > +void intel_mmio_unmap_pci_bar(struct intel_mmio_data *mmio_data);
> > void intel_mmio_use_dump_file(struct intel_mmio_data *mmio_data, char *file);
> > +void intel_mmio_unmap_dump_file(struct intel_mmio_data *mmio_data);
> >
> > int intel_register_access_init(struct intel_mmio_data *mmio_data,
> > struct pci_device *pci_dev, int safe, int fd);
> > diff --git a/lib/intel_mmio.c b/lib/intel_mmio.c
> > index 667a69f5aa..cb8f9db2e5 100644
> > --- a/lib/intel_mmio.c
> > +++ b/lib/intel_mmio.c
> > @@ -82,6 +82,8 @@ void *igt_global_mmio;
> > * Sets also up mmio_data->igt_mmio to point at the data contained
> > * in @file. This allows the same code to get reused for dumping and decoding
> > * from running hardware as from register dumps.
> > + *
> > + * Users are expected to call intel_mmio_unmap_dump_file() after use.
> > */
> > void
> > intel_mmio_use_dump_file(struct intel_mmio_data *mmio_data, char *file)
> > @@ -99,11 +101,29 @@ intel_mmio_use_dump_file(struct intel_mmio_data *mmio_data, char *file)
>
> imho at beginning of this function there should be check
> that igt_global_mmio == NULL, and the same check should be at
> other init functions.
No, what I think needs to be fixed are users who are still using
igt_global_mmio while they should use the value they passed as the mmio
argument to intel_mmio_use_*() or intel_register_access_init(), and that
global variable should be dropped. But first of all, that's not related to
the issue this patch is trying to fix, then out of scope of this patch.
> Looks like we cannot mmap two different pcie cards at the same
> time with this lib.
We can, if we just ignore that depreciated global variable, I believe.
> > igt_fail_on_f(mmio_data->igt_mmio == MAP_FAILED,
> > "Couldn't mmap %s\n", file);
> >
> > + mmio_data->mmio_size = st.st_size;
> > igt_global_mmio = mmio_data->igt_mmio;
> >
> > close(fd);
> > }
> >
> > +/**
> > + * intel_mmio_unmap_dump_file:
> > + * @mmio_data: mmio structure for IO operations
> > + *
> > + * Unmaps a dump file mmapped with intel_mmio_use_dump_file()
> > + */
> > +void intel_mmio_unmap_dump_file(struct intel_mmio_data *mmio_data)
> > +{
> > + if (igt_warn_on_f(!mmio_data->mmio_size || mmio_data->dev,
> > + "test bug: argument doesn't point to struct intel_mmio_data initialized with intel_mmio_use_dump_file()\n"))
>
> Please shorten text for warning, something like: arg was not
> inialized with ...
OK
> Please also add check for null at global var.
>
> > + return;
> > +
> > + igt_global_mmio = NULL;
> > + igt_debug_on(munmap(mmio_data->igt_mmio, mmio_data->mmio_size) < 0);
> > + mmio_data->mmio_size = 0;
> > +}
> > +
> > /**
> > * intel_mmio_use_pci_bar:
> > * @mmio_data: mmio structure for IO operations
> > @@ -112,12 +132,14 @@ intel_mmio_use_dump_file(struct intel_mmio_data *mmio_data, char *file)
> > * Fill a mmio_data stucture with igt_mmio to point at the mmio bar.
> > *
> > * @pci_dev can be obtained from intel_get_pci_device().
> > + *
> > + * Users are expected to call intel_mmio_unmap_pci_bar() after use.
> > */
> > void
> > intel_mmio_use_pci_bar(struct intel_mmio_data *mmio_data, struct pci_device *pci_dev)
> > {
> > uint32_t devid, gen;
> > - int mmio_bar, mmio_size;
> > + int mmio_bar;
>
> Please use this local var and assign it to struct only after
> succesfull initialization.
OK
>
> > int error;
> >
> > memset(mmio_data, 0, sizeof(struct intel_mmio_data));
> > @@ -129,22 +151,42 @@ intel_mmio_use_pci_bar(struct intel_mmio_data *mmio_data, struct pci_device *pci
> >
> > gen = intel_gen(devid);
> > if (gen < 3)
> > - mmio_size = 512*1024;
> > + mmio_data->mmio_size = 512*1024;
> > else if (gen < 5)
> > - mmio_size = 512*1024;
> > + mmio_data->mmio_size = 512*1024;
>
> Both places uses the same number 512*1024, please make it one
> if check: if (gen < 5)
>
> Or maybe it is an error for gen < 3 ?
Out of scope.
>
> > else
> > - mmio_size = 2*1024*1024;
> > + mmio_data->mmio_size = 2*1024*1024;
> >
> > error = pci_device_map_range (pci_dev,
> > pci_dev->regions[mmio_bar].base_addr,
> > - mmio_size,
> > + mmio_data->mmio_size,
> > PCI_DEV_MAP_FLAG_WRITABLE,
> > &mmio_data->igt_mmio);
> >
> > - igt_global_mmio = mmio_data->igt_mmio;
> > -
> > igt_fail_on_f(error != 0,
> > "Couldn't map MMIO region\n");
> > +
> > + mmio_data->dev = pci_dev;
> > + igt_global_mmio = mmio_data->igt_mmio;
> > +}
> > +
> > +/**
> > + * intel_mmio_unmap_pci_bar:
> > + * @mmio_data: mmio structure for IO operations
> > + *
> > + * Unmaps a PCI BAR region mmapped with intel_mmio_use_pci_bar()
> > + */
> > +void intel_mmio_unmap_pci_bar(struct intel_mmio_data *mmio_data)
> > +{
> > + if (igt_warn_on_f(!mmio_data->dev,
> > + "test bug: argument doesn't point to struct intel_mmio_data initialized with intel_mmio_use_pci_bar()\n"))
>
> Same here, please shorten this warn.
>
> > + return;
> > +
> > + igt_global_mmio = NULL;
> > + igt_debug_on(pci_device_unmap_range(mmio_data->dev,
> > + mmio_data->igt_mmio, mmio_data->mmio_size) < 0);
> > + mmio_data->dev = NULL;
> > + mmio_data->mmio_size = 0;
> > }
> >
> > static void
> > @@ -166,6 +208,8 @@ release_forcewake_lock(int fd)
> > * It also initializes mmio_data->igt_mmio like intel_mmio_use_pci_bar().
> > *
> > * @pci_dev can be obtained from intel_get_pci_device().
> > + *
> > + * Users are expected to call intel_register_access_fini() after use.
> > */
> > int
> > intel_register_access_init(struct intel_mmio_data *mmio_data, struct pci_device *pci_dev, int safe, int fd)
> > @@ -222,8 +266,15 @@ int intel_register_access_needs_fakewake(struct intel_mmio_data *mmio_data)
> > void
> > intel_register_access_fini(struct intel_mmio_data *mmio_data)
> > {
> > - if (mmio_data->key && intel_register_access_needs_wake(mmio_data))
> > + if (igt_warn_on_f(!mmio_data->key,
> > + "test bug: argument doesn't point to struct intel_mmio_data initialized with intel_register_access_init()\n"))
>
> Same here, please shorten this warn.
>
> Btw, in this lib error condition for key is -1, so maybe this
> should also be cheked ?
No, key == -1 means forcewake is not supported and shouldn't be used but it
still means the structure was initialized by intel_register_access_init().
However, thanks for this question, since now I can see 0, though unlikely, can
be a valid key.
AFAICT, 0000 is not a valid PCI device ID, then mmio->device_id is a much
better candidate than ->key for this check -- I'll switch to it in v2 of this
patch.
> > + return;
> > +
> > + if (intel_register_access_needs_wake(mmio_data))
> > release_forcewake_lock(mmio_data->key);
> > + mmio_data->key = 0;
> > +
> > + intel_mmio_unmap_pci_bar(mmio_data);
> > }
> >
> > /**
> >
>
> Please correct desciption of global var igt_global_mmio, there
> is one more method for initialize it: intel_mmio_use_pci_bar
> as you wrote on irc.
Out of scope.
Thanks,
Janusz
>
> Regards,
> Kamil Konieczny
>
>
More information about the Intel-gfx
mailing list