On Mon, Aug 12, 2019 at 11:42 AM Marc Koderer <marc@xxxxxxxxxxx> wrote:
>
> There is nearly no difference between both implemenations.
> ceph_kvmalloc existed before kvmalloc which makes me think it's
> a leftover.
>
> Signed-off-by: Marc Koderer <marc@xxxxxxxxxxx>
> ---
> net/ceph/buffer.c | 3 +--
> net/ceph/ceph_common.c | 11 -----------
> net/ceph/crypto.c | 2 +-
> net/ceph/messenger.c | 2 +-
> 4 files changed, 3 insertions(+), 15 deletions(-)
>
> diff --git a/net/ceph/buffer.c b/net/ceph/buffer.c
> index 5622763ad402..6ca273d2246a 100644
> --- a/net/ceph/buffer.c
> +++ b/net/ceph/buffer.c
> @@ -7,7 +7,6 @@
>
> #include <linux/ceph/buffer.h>
> #include <linux/ceph/decode.h>
> -#include <linux/ceph/libceph.h> /* for ceph_kvmalloc */
>
> struct ceph_buffer *ceph_buffer_new(size_t len, gfp_t gfp)
> {
> @@ -17,7 +16,7 @@ struct ceph_buffer *ceph_buffer_new(size_t len, gfp_t gfp)
> if (!b)
> return NULL;
>
> - b->vec.iov_base = ceph_kvmalloc(len, gfp);
> + b->vec.iov_base = kvmalloc(len, gfp);
> if (!b->vec.iov_base) {
> kfree(b);
> return NULL;
> diff --git a/net/ceph/ceph_common.c b/net/ceph/ceph_common.c
> index 4eeea4d5c3ef..6c1769a815af 100644
> --- a/net/ceph/ceph_common.c
> +++ b/net/ceph/ceph_common.c
> @@ -185,17 +185,6 @@ int ceph_compare_options(struct ceph_options *new_opt,
> }
> EXPORT_SYMBOL(ceph_compare_options);
>
> -void *ceph_kvmalloc(size_t size, gfp_t flags)
> -{
> - if (size <= (PAGE_SIZE << PAGE_ALLOC_COSTLY_ORDER)) {
> - void *ptr = kmalloc(size, flags | __GFP_NOWARN);
> - if (ptr)
> - return ptr;
> - }
> -
> - return __vmalloc(size, flags, PAGE_KERNEL);
> -}
> -
>
> static int parse_fsid(const char *str, struct ceph_fsid *fsid)
> {
> diff --git a/net/ceph/crypto.c b/net/ceph/crypto.c
> index 5d6724cee38f..a9deead1e4ff 100644
> --- a/net/ceph/crypto.c
> +++ b/net/ceph/crypto.c
> @@ -144,7 +144,7 @@ void ceph_crypto_key_destroy(struct ceph_crypto_key *key)
> static const u8 *aes_iv = (u8 *)CEPH_AES_IV;
>
> /*
> - * Should be used for buffers allocated with ceph_kvmalloc().
> + * Should be used for buffers allocated with kvmalloc().
> * Currently these are encrypt out-buffer (ceph_buffer) and decrypt
> * in-buffer (msg front).
> *
> diff --git a/net/ceph/messenger.c b/net/ceph/messenger.c
> index 962f521c863e..f1f2fcc6f780 100644
> --- a/net/ceph/messenger.c
> +++ b/net/ceph/messenger.c
> @@ -3334,7 +3334,7 @@ struct ceph_msg *ceph_msg_new2(int type, int front_len, int max_data_items,
>
> /* front */
> if (front_len) {
> - m->front.iov_base = ceph_kvmalloc(front_len, flags);
> + m->front.iov_base = kvmalloc(front_len, flags);
> if (m->front.iov_base == NULL) {
> dout("ceph_msg_new can't allocate %d bytes\n",
> front_len);
Hi Marc,
I'm working on a patch for https://tracker.ceph.com/issues/40481 which
changes ceph_kvmalloc() to properly deal with non-GFP_KERNEL contexts.
We can't switch to kvmalloc() because it doesn't actually fall back to
vmalloc() for GFP_NOFS or GFP_NOIO.
Thanks,
Ilya
