-----"Jason Gunthorpe" <jgg@xxxxxxxx> wrote: ----- >To: "Bernard Metzler" <BMT@xxxxxxxxxxxxxx> >From: "Jason Gunthorpe" <jgg@xxxxxxxx> >Date: 05/07/2019 07:09PM >Cc: "Leon Romanovsky" <leon@xxxxxxxxxx>, linux-rdma@xxxxxxxxxxxxxxx >Subject: Re: [PATCH v8 02/12] SIW main include file > >On Tue, May 07, 2019 at 03:54:45PM +0000, Bernard Metzler wrote: >> >> >To: "Bernard Metzler" <BMT@xxxxxxxxxxxxxx> >> >From: "Leon Romanovsky" <leon@xxxxxxxxxx> >> >Date: 05/05/2019 07:10PM >> >Cc: linux-rdma@xxxxxxxxxxxxxxx, "Bernard Metzler" >> ><bmt@xxxxxxxxxxxxxxxxxxx> >> >Subject: Re: [PATCH v8 02/12] SIW main include file >> > >> >On Sun, May 05, 2019 at 04:54:50PM +0000, Bernard Metzler wrote: >> >> >> >> >To: "Bernard Metzler" <bmt@xxxxxxxxxxxxxx> >> >> >From: "Leon Romanovsky" <leon@xxxxxxxxxx> >> >> >Date: 04/28/2019 01:07PM >> >> >Cc: linux-rdma@xxxxxxxxxxxxxxx, "Bernard Metzler" >> >> ><bmt@xxxxxxxxxxxxxxxxxxx> >> >> >Subject: Re: [PATCH v8 02/12] SIW main include file >> >> > >> >> >On Fri, Apr 26, 2019 at 03:18:42PM +0200, Bernard Metzler >wrote: >> >> >> From: Bernard Metzler <bmt@xxxxxxxxxxxxxxxxxxx> >> >> >> >> >> >> Signed-off-by: Bernard Metzler <bmt@xxxxxxxxxxxxxx> >> >> >> drivers/infiniband/sw/siw/siw.h | 733 >> >> >++++++++++++++++++++++++++++++++ >> >> >> 1 file changed, 733 insertions(+) >> >> >> create mode 100644 drivers/infiniband/sw/siw/siw.h >> >> >> >> >> >> diff --git a/drivers/infiniband/sw/siw/siw.h >> >> >b/drivers/infiniband/sw/siw/siw.h >> >> >> new file mode 100644 >> >> >> index 000000000000..9a3c2abbd858 >> >> >> +++ b/drivers/infiniband/sw/siw/siw.h >> >> >> @@ -0,0 +1,733 @@ >> >> >> +/* SPDX-License-Identifier: GPL-2.0 or BSD-3-Clause */ >> >> >> + >> >> >> +/* Authors: Bernard Metzler <bmt@xxxxxxxxxxxxxx> */ >> >> >> +/* Copyright (c) 2008-2019, IBM Corporation */ >> >> >> + >> >> >> +#ifndef _SIW_H >> >> >> +#define _SIW_H >> >> >> + >> >> >> +#include <linux/idr.h> >> >> >> +#include <rdma/ib_verbs.h> >> >> >> +#include <linux/socket.h> >> >> >> +#include <linux/skbuff.h> >> >> >> +#include <linux/in.h> >> >> >> +#include <linux/fs.h> >> >> >> +#include <linux/netdevice.h> >> >> >> +#include <crypto/hash.h> >> >> >> +#include <linux/resource.h> /* MLOCK_LIMIT */ >> >> >> +#include <linux/module.h> >> >> >> +#include <linux/version.h> >> >> >> +#include <linux/llist.h> >> >> >> +#include <linux/mm.h> >> >> >> +#include <linux/sched/signal.h> >> >> >> + >> >> >> +#include <rdma/siw_user.h> >> >> >> +#include "iwarp.h" >> >> >> + >> >> >> +/* driver debugging enabled */ >> >> >> +#define DEBUG >> >> > >> >> >I clearly remember that we asked to remove this. >> >> >> >> Absolutely. Sorry, it sneaked in again since I did some >> >> debugging. Will remove... >> >> > >> >> >> + spinlock_t lock; >> >> >> + >> >> >> + /* object management */ >> >> >> + struct idr qp_idr; >> >> >> + struct idr mem_idr; >> >> > >> >> >Why IDR and not XArray? >> >> >> >> Memory access keys and QP IDs are generated as random >> >> numbers, since both are exposed to the application. >> >> Since XArray is not designed for sparsely distributed >> >> id ranges, I am still in favor of IDR for these two >> >> resources. > >IDR and xarray have identical underlying storage so this is nonsense > >No new idr's or radix tree users will be accepted into rdma.... Use >xarray > Sounds good to me! I just came across that introductory video from Matthew, where he explicitly stated that xarray will be not very efficient if the indices are not densely clustered. But maybe this is all far beyond the 24bits of index space a memory key is in. So let me drop that IDR thing completely, while handling randomized 24 bit memory keys. Thanks Bernard
