diff mbox series

[net-next,v22,04/13] netdev: netdevice devmem allocator

Message ID 20240825041511.324452-5-almasrymina@google.com (mailing list archive)
State New
Headers show
Series Device Memory TCP | expand

Commit Message

Mina Almasry Aug. 25, 2024, 4:15 a.m. UTC 
Implement netdev devmem allocator. The allocator takes a given struct
netdev_dmabuf_binding as input and allocates net_iov from that
binding.

The allocation simply delegates to the binding's genpool for the
allocation logic and wraps the returned memory region in a net_iov
struct.

Signed-off-by: Willem de Bruijn <willemb@google.com>
Signed-off-by: Kaiyuan Zhang <kaiyuanz@google.com>
Signed-off-by: Mina Almasry <almasrymina@google.com>
Reviewed-by: Pavel Begunkov <asml.silence@gmail.com>

---

v20:
- Removed dma_addr field in dmabuf_genpool_chunk_owner not used in this
  patch (moved to later patch where it's used).

v19:
- Don't reset dma_addr on allocation/free (Jakub)

v17:
- Don't acquire a binding ref for every allocation (Jakub).

v11:
- Fix extraneous inline directive (Paolo)

v8:
- Rename netdev_dmabuf_binding -> net_devmem_dmabuf_binding to avoid
  patch-by-patch build error.
- Move niov->pp_magic/pp/pp_ref_counter usage to later patch to avoid
  patch-by-patch build error.

v7:
- netdev_ -> net_devmem_* naming (Yunsheng).

v6:
- Add comment on net_iov_dma_addr to explain why we don't use
  niov->dma_addr (Pavel)
- Refactor new functions into net/core/devmem.c (Pavel)

v1:
- Rename devmem -> dmabuf (David).

---
 include/net/devmem.h | 13 +++++++++++++
 include/net/netmem.h | 17 +++++++++++++++++
 net/core/devmem.c    | 38 ++++++++++++++++++++++++++++++++++++++
 3 files changed, 68 insertions(+)

Comments

Jakub Kicinski Aug. 28, 2024, 2:15 a.m. UTC | #1 
On Sun, 25 Aug 2024 04:15:02 +0000 Mina Almasry wrote:
> +void net_devmem_free_dmabuf(struct net_iov *niov)
> +{
> +	struct net_devmem_dmabuf_binding *binding = net_iov_binding(niov);
> +	unsigned long dma_addr = net_devmem_get_dma_addr(niov);
> +
> +	if (gen_pool_has_addr(binding->chunk_pool, dma_addr, PAGE_SIZE))
> +		gen_pool_free(binding->chunk_pool, dma_addr, PAGE_SIZE);

Is the check necessary for correctness? Should it perhaps be a WARN
under DEBUG_NET instead? The rest LGTM:

Reviewed-by: Jakub Kicinski <kuba@kernel.org>
Mina Almasry Aug. 28, 2024, 7:20 a.m. UTC | #2 
On Tue, Aug 27, 2024 at 7:15 PM Jakub Kicinski <kuba@kernel.org> wrote:
>
> On Sun, 25 Aug 2024 04:15:02 +0000 Mina Almasry wrote:
> > +void net_devmem_free_dmabuf(struct net_iov *niov)
> > +{
> > +     struct net_devmem_dmabuf_binding *binding = net_iov_binding(niov);
> > +     unsigned long dma_addr = net_devmem_get_dma_addr(niov);
> > +
> > +     if (gen_pool_has_addr(binding->chunk_pool, dma_addr, PAGE_SIZE))
> > +             gen_pool_free(binding->chunk_pool, dma_addr, PAGE_SIZE);
>
> Is the check necessary for correctness? Should it perhaps be a WARN
> under DEBUG_NET instead? The rest LGTM:
>

Not really necessary for correctness per se, but if we try to free a
dma_addr that is not in a gen_pool (due to some other bug in the
code), then gen_pool_free ends up BUG_ON, crashing the kernel.

Arguably gen_pool_free should not BUG_ON, but I think that's an old
API, and existing call sites have worked around the BUG_ON by doing a
gen_pool_has_addr check like I do here, for example kernel/dma/pool.c.
So I did not seek to change this established behavior.

I think WARN seems fine to me, but maybe not under DEBUG_NET. I don't
want production code crashing due to this error, if it's OK with you.

Unless I hear otherwise I'll add a WARN without debug here.

> Reviewed-by: Jakub Kicinski <kuba@kernel.org>
>

Thanks!
Jakub Kicinski Aug. 28, 2024, 6:43 p.m. UTC | #3 
On Wed, 28 Aug 2024 00:20:23 -0700 Mina Almasry wrote:
> > On Sun, 25 Aug 2024 04:15:02 +0000 Mina Almasry wrote:  
> > > +void net_devmem_free_dmabuf(struct net_iov *niov)
> > > +{
> > > +     struct net_devmem_dmabuf_binding *binding = net_iov_binding(niov);
> > > +     unsigned long dma_addr = net_devmem_get_dma_addr(niov);
> > > +
> > > +     if (gen_pool_has_addr(binding->chunk_pool, dma_addr, PAGE_SIZE))
> > > +             gen_pool_free(binding->chunk_pool, dma_addr, PAGE_SIZE);  
> >
> > Is the check necessary for correctness? Should it perhaps be a WARN
> > under DEBUG_NET instead? The rest LGTM:
> >  
> 
> Not really necessary for correctness per se, but if we try to free a
> dma_addr that is not in a gen_pool (due to some other bug in the
> code), then gen_pool_free ends up BUG_ON, crashing the kernel.
> 
> Arguably gen_pool_free should not BUG_ON, but I think that's an old
> API, and existing call sites have worked around the BUG_ON by doing a
> gen_pool_has_addr check like I do here, for example kernel/dma/pool.c.
> So I did not seek to change this established behavior.
> 
> I think WARN seems fine to me, but maybe not under DEBUG_NET. I don't
> want production code crashing due to this error, if it's OK with you.
> 
> Unless I hear otherwise I'll add a WARN without debug here.

WARN makes sense, I didn't know about the BUG_ON() hiding inside
gen_pool :(
diff mbox series

Patch

diff --git a/include/net/devmem.h b/include/net/devmem.h
index 02925d897aa6..7b856a3540a9 100644
--- a/include/net/devmem.h
+++ b/include/net/devmem.h
@@ -72,7 +72,20 @@  void net_devmem_unbind_dmabuf(struct net_devmem_dmabuf_binding *binding);
 int net_devmem_bind_dmabuf_to_queue(struct net_device *dev, u32 rxq_idx,
 				    struct net_devmem_dmabuf_binding *binding);
 void dev_dmabuf_uninstall(struct net_device *dev);
+struct net_iov *
+net_devmem_alloc_dmabuf(struct net_devmem_dmabuf_binding *binding);
+void net_devmem_free_dmabuf(struct net_iov *ppiov);
 #else
+static inline struct net_iov *
+net_devmem_alloc_dmabuf(struct net_devmem_dmabuf_binding *binding)
+{
+	return NULL;
+}
+
+static inline void net_devmem_free_dmabuf(struct net_iov *ppiov)
+{
+}
+
 static inline void
 __net_devmem_dmabuf_binding_free(struct net_devmem_dmabuf_binding *binding)
 {
diff --git a/include/net/netmem.h b/include/net/netmem.h
index 41e96c2f94b5..0fbc0999091a 100644
--- a/include/net/netmem.h
+++ b/include/net/netmem.h
@@ -16,6 +16,23 @@  struct net_iov {
 	struct dmabuf_genpool_chunk_owner *owner;
 };
 
+static inline struct dmabuf_genpool_chunk_owner *
+net_iov_owner(const struct net_iov *niov)
+{
+	return niov->owner;
+}
+
+static inline unsigned int net_iov_idx(const struct net_iov *niov)
+{
+	return niov - net_iov_owner(niov)->niovs;
+}
+
+static inline struct net_devmem_dmabuf_binding *
+net_iov_binding(const struct net_iov *niov)
+{
+	return net_iov_owner(niov)->binding;
+}
+
 /* netmem */
 
 /**
diff --git a/net/core/devmem.c b/net/core/devmem.c
index b2b271a921c1..d636a8b34215 100644
--- a/net/core/devmem.c
+++ b/net/core/devmem.c
@@ -35,6 +35,14 @@  static void net_devmem_dmabuf_free_chunk_owner(struct gen_pool *genpool,
 	kfree(owner);
 }
 
+static dma_addr_t net_devmem_get_dma_addr(const struct net_iov *niov)
+{
+	struct dmabuf_genpool_chunk_owner *owner = net_iov_owner(niov);
+
+	return owner->base_dma_addr +
+	       ((dma_addr_t)net_iov_idx(niov) << PAGE_SHIFT);
+}
+
 void __net_devmem_dmabuf_binding_free(struct net_devmem_dmabuf_binding *binding)
 {
 	size_t size, avail;
@@ -57,6 +65,36 @@  void __net_devmem_dmabuf_binding_free(struct net_devmem_dmabuf_binding *binding)
 	kfree(binding);
 }
 
+struct net_iov *
+net_devmem_alloc_dmabuf(struct net_devmem_dmabuf_binding *binding)
+{
+	struct dmabuf_genpool_chunk_owner *owner;
+	unsigned long dma_addr;
+	struct net_iov *niov;
+	ssize_t offset;
+	ssize_t index;
+
+	dma_addr = gen_pool_alloc_owner(binding->chunk_pool, PAGE_SIZE,
+					(void **)&owner);
+	if (!dma_addr)
+		return NULL;
+
+	offset = dma_addr - owner->base_dma_addr;
+	index = offset / PAGE_SIZE;
+	niov = &owner->niovs[index];
+
+	return niov;
+}
+
+void net_devmem_free_dmabuf(struct net_iov *niov)
+{
+	struct net_devmem_dmabuf_binding *binding = net_iov_binding(niov);
+	unsigned long dma_addr = net_devmem_get_dma_addr(niov);
+
+	if (gen_pool_has_addr(binding->chunk_pool, dma_addr, PAGE_SIZE))
+		gen_pool_free(binding->chunk_pool, dma_addr, PAGE_SIZE);
+}
+
 void net_devmem_unbind_dmabuf(struct net_devmem_dmabuf_binding *binding)
 {
 	struct netdev_rx_queue *rxq;