| lists.openwall.net | lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC | |
|
Open Source and information security mailing list archives
| ||
|
Date: Mon, 7 Jun 2021 13:23:35 +0200
From: Greg KH <gregkh@...uxfoundation.org>
To: Leon Romanovsky <leon@...nel.org>
Cc: Doug Ledford <dledford@...hat.com>,
Jason Gunthorpe <jgg@...dia.com>,
Kees Cook <keescook@...omium.org>,
Nathan Chancellor <nathan@...nel.org>,
Adit Ranadive <aditr@...are.com>,
Ariel Elior <aelior@...vell.com>,
Christian Benvenuti <benve@...co.com>,
clang-built-linux@...glegroups.com,
Dennis Dalessandro <dennis.dalessandro@...nelisnetworks.com>,
Devesh Sharma <devesh.sharma@...adcom.com>,
Gal Pressman <galpress@...zon.com>,
linux-kernel@...r.kernel.org, linux-rdma@...r.kernel.org,
Michal Kalderon <mkalderon@...vell.com>,
Mike Marciniszyn <mike.marciniszyn@...nelisnetworks.com>,
Mustafa Ismail <mustafa.ismail@...el.com>,
Naresh Kumar PBS <nareshkumar.pbs@...adcom.com>,
Nelson Escobar <neescoba@...co.com>,
Nick Desaulniers <ndesaulniers@...gle.com>,
Potnuri Bharat Teja <bharat@...lsio.com>,
Selvin Xavier <selvin.xavier@...adcom.com>,
Shiraz Saleem <shiraz.saleem@...el.com>,
VMware PV-Drivers <pv-drivers@...are.com>,
Yishai Hadas <yishaih@...dia.com>,
Zhu Yanjun <zyjzyj2000@...il.com>
Subject: Re: [PATCH rdma-next v1 02/15] RDMA/core: Replace the ib_port_data
hw_stats pointers with a ib_port pointer
On Mon, Jun 07, 2021 at 02:00:26PM +0300, Leon Romanovsky wrote:
> On Mon, Jun 07, 2021 at 12:23:23PM +0200, Greg KH wrote:
> > On Mon, Jun 07, 2021 at 11:17:27AM +0300, Leon Romanovsky wrote:
> > > From: Jason Gunthorpe <jgg@...dia.com>
> > >
> > > It is much saner to store a pointer to the kobject structure that contains
> > > the cannonical stats pointer than to copy the stats pointers into a public
> > > structure.
> > >
> > > Future patches will require the sysfs pointer for other purposes.
> > >
> > > Signed-off-by: Jason Gunthorpe <jgg@...dia.com>
> > > Signed-off-by: Leon Romanovsky <leonro@...dia.com>
> > > ---
> > > drivers/infiniband/core/core_priv.h | 1 +
> > > drivers/infiniband/core/nldev.c | 8 ++------
> > > drivers/infiniband/core/sysfs.c | 14 +++++++++++---
> > > include/rdma/ib_verbs.h | 3 ++-
> > > 4 files changed, 16 insertions(+), 10 deletions(-)
> > >
> > > diff --git a/drivers/infiniband/core/core_priv.h b/drivers/infiniband/core/core_priv.h
> > > index 29809dd30041..ec5c2c3db423 100644
> > > --- a/drivers/infiniband/core/core_priv.h
> > > +++ b/drivers/infiniband/core/core_priv.h
> > > @@ -378,6 +378,7 @@ struct net_device *rdma_read_gid_attr_ndev_rcu(const struct ib_gid_attr *attr);
> > >
> > > void ib_free_port_attrs(struct ib_core_device *coredev);
> > > int ib_setup_port_attrs(struct ib_core_device *coredev);
> > > +struct rdma_hw_stats *ib_get_hw_stats_port(struct ib_device *ibdev, u32 port_num);
> > >
> > > int rdma_compatdev_set(u8 enable);
> > >
> > > diff --git a/drivers/infiniband/core/nldev.c b/drivers/infiniband/core/nldev.c
> > > index 01316926cef6..e9b4b2cccaa0 100644
> > > --- a/drivers/infiniband/core/nldev.c
> > > +++ b/drivers/infiniband/core/nldev.c
> > > @@ -2066,7 +2066,8 @@ static int stat_get_doit_default_counter(struct sk_buff *skb,
> > > }
> > >
> > > port = nla_get_u32(tb[RDMA_NLDEV_ATTR_PORT_INDEX]);
> > > - if (!rdma_is_port_valid(device, port)) {
> > > + stats = ib_get_hw_stats_port(device, port);
> > > + if (!stats) {
> > > ret = -EINVAL;
> > > goto err;
> > > }
> > > @@ -2088,11 +2089,6 @@ static int stat_get_doit_default_counter(struct sk_buff *skb,
> > > goto err_msg;
> > > }
> > >
> > > - stats = device->port_data ? device->port_data[port].hw_stats : NULL;
> > > - if (stats == NULL) {
> > > - ret = -EINVAL;
> > > - goto err_msg;
> > > - }
> > > mutex_lock(&stats->lock);
> > >
> > > num_cnts = device->ops.get_hw_stats(device, stats, port, 0);
> > > diff --git a/drivers/infiniband/core/sysfs.c b/drivers/infiniband/core/sysfs.c
> > > index d11ceff2b4e4..b153dee1e0fa 100644
> > > --- a/drivers/infiniband/core/sysfs.c
> > > +++ b/drivers/infiniband/core/sysfs.c
> > > @@ -1031,8 +1031,6 @@ static void setup_hw_stats(struct ib_device *device, struct ib_port *port,
> > > goto err;
> > > port->hw_stats_ag = hsag;
> > > port->hw_stats = stats;
> > > - if (device->port_data)
> > > - device->port_data[port_num].hw_stats = stats;
> > > } else {
> > > struct kobject *kobj = &device->dev.kobj;
> > > ret = sysfs_create_group(kobj, hsag);
> > > @@ -1053,6 +1051,14 @@ static void setup_hw_stats(struct ib_device *device, struct ib_port *port,
> > > kfree(stats);
> > > }
> > >
> > > +struct rdma_hw_stats *ib_get_hw_stats_port(struct ib_device *ibdev,
> > > + u32 port_num)
> > > +{
> > > + if (!ibdev->port_data || !rdma_is_port_valid(ibdev, port_num))
> > > + return NULL;
> > > + return ibdev->port_data[port_num].sysfs->hw_stats;
> > > +}
> > > +
> > > static int add_port(struct ib_core_device *coredev, int port_num)
> > > {
> > > struct ib_device *device = rdma_device_to_ibdev(&coredev->dev);
> > > @@ -1171,6 +1177,8 @@ static int add_port(struct ib_core_device *coredev, int port_num)
> > > setup_hw_stats(device, p, port_num);
> > >
> > > list_add_tail(&p->kobj.entry, &coredev->port_list);
> > > + if (device->port_data && is_full_dev)
> > > + device->port_data[port_num].sysfs = p;
> >
> > You are saving off a pointer to a reference counted structure without
> > incrementing the reference count on it? That's brave, and really wrong.
>
> This is done to ensure that device->port_data[port_num].sysfs is not set
> before "p" is properly configured. From reference point of view "sysfs"
> is equal to "p" and the latter already initialized.
That's fine, but again, you are saving a pointer to a reference counted
object, and then later using it, without ever having properly
incremented the reference count to assure that it really is going to
stick around.
thanks,
greg k-h
Powered by blists - more mailing lists