Re: [PATCH 1/1] mm: change inlined allocation helpers to account at the call site

[Suren Baghdasaryan <surenb@google.com>]

On Thu, Apr 4, 2024 at 10:04 AM Matthew Wilcox <willy@infradead.org> wrote:
>
> On Thu, Apr 04, 2024 at 09:54:04AM -0700, Suren Baghdasaryan wrote:
> > +++ b/include/linux/dma-fence-chain.h
> > @@ -86,10 +86,7 @@ dma_fence_chain_contained(struct dma_fence *fence)
> >   *
> >   * Returns a new struct dma_fence_chain object or NULL on failure.
> >   */
> > -static inline struct dma_fence_chain *dma_fence_chain_alloc(void)
> > -{
> > -     return kmalloc(sizeof(struct dma_fence_chain), GFP_KERNEL);
> > -};
> > +#define dma_fence_chain_alloc()      kmalloc(sizeof(struct dma_fence_chain), GFP_KERNEL)
>
> You've removed some typesafety here.  Before, if I wrote:
>
>         struct page *page = dma_fence_chain_alloc();
>
> the compiler would warn me that I've done something stupid.  Now it
> can't tell.  Suggest perhaps:
>
> #define dma_fence_chain_alloc()                                           \
>         (struct dma_fence_chain *)kmalloc(sizeof(struct dma_fence_chain), \
>                                                 GFP_KERNEL)
>
> but maybe there's a better way of doing that.  There are a few other
> occurrences of the same problem in this monster patch.

Got your point.

>
> > +++ b/include/linux/hid_bpf.h
> > @@ -149,10 +149,7 @@ static inline int hid_bpf_connect_device(struct hid_device *hdev) { return 0; }
> >  static inline void hid_bpf_disconnect_device(struct hid_device *hdev) {}
> >  static inline void hid_bpf_destroy_device(struct hid_device *hid) {}
> >  static inline void hid_bpf_device_init(struct hid_device *hid) {}
> > -static inline u8 *call_hid_bpf_rdesc_fixup(struct hid_device *hdev, u8 *rdesc, unsigned int *size)
> > -{
> > -     return kmemdup(rdesc, *size, GFP_KERNEL);
> > -}
> > +#define call_hid_bpf_rdesc_fixup(_hdev, _rdesc, _size) kmemdup(_rdesc, *(_size), GFP_KERNEL)
>
> here
>
> > -static inline handle_t *jbd2_alloc_handle(gfp_t gfp_flags)
> > -{
> > -     return kmem_cache_zalloc(jbd2_handle_cache, gfp_flags);
> > -}
> > +#define jbd2_alloc_handle(_gfp_flags)        kmem_cache_zalloc(jbd2_handle_cache, _gfp_flags)
>
> here
>
> > +++ b/include/linux/skmsg.h
> > @@ -410,11 +410,8 @@ void sk_psock_stop_verdict(struct sock *sk, struct sk_psock *psock);
> >  int sk_psock_msg_verdict(struct sock *sk, struct sk_psock *psock,
> >                        struct sk_msg *msg);
> >
> > -static inline struct sk_psock_link *sk_psock_init_link(void)
> > -{
> > -     return kzalloc(sizeof(struct sk_psock_link),
> > -                    GFP_ATOMIC | __GFP_NOWARN);
> > -}
> > +#define sk_psock_init_link() \
> > +             kzalloc(sizeof(struct sk_psock_link), GFP_ATOMIC | __GFP_NOWARN)
>
> here
>
> ... I kind of gave up at this point.  You'll want to audit for yourself
> anyway ;-)

Yes, I'll go over it and will make the required changes. Thanks for
looking into it!
Suren.