> From: linux-rdma-owner@xxxxxxxxxxxxxxx <linux-rdma-
> owner@xxxxxxxxxxxxxxx> On Behalf Of Gal Pressman
> Sent: Thursday, June 27, 2019 5:46 PM
> To: Michal Kalderon <mkalderon@xxxxxxxxxxx>; jgg@xxxxxxxx;
> dledford@xxxxxxxxxx; leon@xxxxxxxxxx; sleybo@xxxxxxxxxx; Ariel Elior
> <aelior@xxxxxxxxxxx>
> Cc: linux-rdma@xxxxxxxxxxxxxxx
> Subject: Re: [RFC rdma 1/3] RDMA/core: Create a common mmap function
>
> On 27/06/2019 16:58, Michal Kalderon wrote:
> > +/*
> > + * Note this locking scheme cannot support removal of entries, except
> > +during
> > + * ucontext destruction when the core code guarentees no concurrency.
> > + */
> > +u64 rdma_user_mmap_entry_insert(struct ib_ucontext *ucontext, void
> *obj,
> > + u64 address, u64 length, u8 mmap_flag) {
> > + struct rdma_user_mmap_entry *entry;
> > + int err;
> > +
> > + entry = kmalloc(sizeof(*entry), GFP_KERNEL);
> > + if (!entry)
> > + return RDMA_USER_MMAP_INVALID;
> > +
> > + entry->obj = obj;
> > + entry->address = address;
> > + entry->length = length;
> > + entry->mmap_flag = mmap_flag;
> > +
> > + xa_lock(&ucontext->mmap_xa);
> > + entry->mmap_page = ucontext->mmap_xa_page;
> > + ucontext->mmap_xa_page += DIV_ROUND_UP(length, PAGE_SIZE);
>
> Hi Michal,
> I fixed this part to handle mmap_xa_page overflow:
> 7a5834e456f7 ("RDMA/efa: Handle mmap insertions overflow")
>
Thanks, will take the patch
> > + err = __xa_insert(&ucontext->mmap_xa, entry->mmap_page,
> entry,
> > + GFP_KERNEL);
> > + xa_unlock(&ucontext->mmap_xa);
> > + if (err) {
> > + kfree(entry);
> > + return RDMA_USER_MMAP_INVALID;
> > + }
> > +
> > + ibdev_dbg(ucontext->device,
> > + "mmap: obj[0x%p] addr[%#llx], len[%#llx], key[%#llx]
> inserted\n",
> > + entry->obj, entry->address, entry->length,
> > + rdma_user_mmap_get_key(entry));
> > +
> > + return rdma_user_mmap_get_key(entry); }
> > +EXPORT_SYMBOL(rdma_user_mmap_entry_insert);
