[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Message-ID: <20250217100614.3043620-2-pasic@linux.ibm.com>
Date: Mon, 17 Feb 2025 11:06:13 +0100
From: Halil Pasic <pasic@...ux.ibm.com>
To: Eric Farman <farman@...ux.ibm.com>,
Matthew Rosato <mjrosato@...ux.ibm.com>,
Halil Pasic <pasic@...ux.ibm.com>,
Vineeth Vijayan <vneethv@...ux.ibm.com>,
Peter Oberparleiter <oberpar@...ux.ibm.com>,
Heiko Carstens <hca@...ux.ibm.com>, Vasily Gorbik <gor@...ux.ibm.com>,
Alexander Gordeev <agordeev@...ux.ibm.com>,
Christian Borntraeger <borntraeger@...ux.ibm.com>,
Sven Schnelle <svens@...ux.ibm.com>,
Tony Krowiak <akrowiak@...ux.ibm.com>,
Jason Herne <jjherne@...ux.ibm.com>,
Harald Freudenberger <freude@...ux.ibm.com>,
Holger Dengler <dengler@...ux.ibm.com>, linux-s390@...r.kernel.org,
kvm@...r.kernel.org, linux-kernel@...r.kernel.org
Cc: "Gustavo A. R. Silva" <gustavo@...eddedor.com>,
Thorsten Blum <thorsten.blum@...ux.dev>
Subject: [PATCH 1/2] s390/vfio-ap: make mdev_types not look like a fake flex array
The vfio-ap driver and the vfio parent device provided by it
(matrix_dev) support just a single mdev_type, and this is not likely to
change any time soon. Despite that matrix_dev->mdev_types started out
as a C99 flexible array presumably as a typo, and since the typo messed
up the allocation, commit e2c8cee9f489 ("s390/vfio-ap: Fix memory
allocation for mdev_types array") changed it to an array of size 1. And
to make things worse mdev_types happens to be the last member of struct
ap_matrix_dev.
Now the problem with that is that before C99 the usual way to get
something similar to a flexible array member was to use a trailing array of
size 0 or 1. This is what I called fake flex array. For a while now the
community is trying to get rid of fake flex arrays. And while mdev_types
is not a fake flex array but an array of size one (to match the mdev
interfaces nicer), it can easily be and was mistaken for a fake flex
array.
So, let us make mdev_types a pointer to struct mdev_type and pass in the
address of that pointer as the 4th formal parameter of
mdev_register_parent().
Signed-off-by: Halil Pasic <pasic@...ux.ibm.com>
Reviewed-by: Anthony Krowiak <akrowiak@...ux.ibm.com>
Reviewed-by: Matthew Rosato <mjrosato@...ux.ibm.com>
Reviewed-by: Eric Farman <farman@...ux.ibm.com>
Tested-by: Anthony Krowiak <akrowiak@...ux.ibm.com>
---
I've also considered switching up the order in which the members
mdev_types and mdev_type are defined in struct ap_matrix_dev but
decided against that because that could look to somebody like
well known mistake that can be made when using fake flex arrays.
---
drivers/s390/crypto/vfio_ap_ops.c | 4 ++--
drivers/s390/crypto/vfio_ap_private.h | 2 +-
2 files changed, 3 insertions(+), 3 deletions(-)
diff --git a/drivers/s390/crypto/vfio_ap_ops.c b/drivers/s390/crypto/vfio_ap_ops.c
index a52c2690933f..5212b3863aff 100644
--- a/drivers/s390/crypto/vfio_ap_ops.c
+++ b/drivers/s390/crypto/vfio_ap_ops.c
@@ -2316,10 +2316,10 @@ int vfio_ap_mdev_register(void)
matrix_dev->mdev_type.sysfs_name = VFIO_AP_MDEV_TYPE_HWVIRT;
matrix_dev->mdev_type.pretty_name = VFIO_AP_MDEV_NAME_HWVIRT;
- matrix_dev->mdev_types[0] = &matrix_dev->mdev_type;
+ matrix_dev->mdev_types = &matrix_dev->mdev_type;
ret = mdev_register_parent(&matrix_dev->parent, &matrix_dev->device,
&vfio_ap_matrix_driver,
- matrix_dev->mdev_types, 1);
+ &matrix_dev->mdev_types, 1);
if (ret)
goto err_driver;
return 0;
diff --git a/drivers/s390/crypto/vfio_ap_private.h b/drivers/s390/crypto/vfio_ap_private.h
index 437a161c8659..9d16321777c8 100644
--- a/drivers/s390/crypto/vfio_ap_private.h
+++ b/drivers/s390/crypto/vfio_ap_private.h
@@ -53,7 +53,7 @@ struct ap_matrix_dev {
struct mutex guests_lock; /* serializes access to each KVM guest */
struct mdev_parent parent;
struct mdev_type mdev_type;
- struct mdev_type *mdev_types[1];
+ struct mdev_type *mdev_types;
};
extern struct ap_matrix_dev *matrix_dev;
--
2.45.2
Powered by blists - more mailing lists