Re: [Xen-devel] [PATCH] xenforeignmemory: work around bug in older privcmd

2018-08-28 Thread Wei Liu 
On Tue, Aug 28, 2018 at 03:12:32PM +0100, Andrew Cooper wrote:
> On 24/08/18 13:16, Paul Durrant wrote:
> > diff --git a/tools/libs/foreignmemory/linux.c 
> > b/tools/libs/foreignmemory/linux.c
> > index 3686cf41e0..0368aa09f4 100644
> > --- a/tools/libs/foreignmemory/linux.c
> > +++ b/tools/libs/foreignmemory/linux.c
> > @@ -53,6 +53,23 @@ int osdep_xenforeignmemory_open(xenforeignmemory_handle 
> > *fmem)
> >  return -1;
> >  }
> >  
> > +/*
> > + * Older versions of privcmd return -EINVAL for unimplemented ioctls
> > + * so we need to probe for the errno to use rather than just using
> > + * the conventional ENOTTY.
> > + */
> > +if ( ioctl(fd, IOCTL_PRIVCMD_UNIMPLEMENTED, NULL) >= 0 )
> > +{
> > +xtl_log(fmem->logger, XTL_ERROR, -1, "xenforeignmemory",
> > +"privcmd ioctl should not be implemented");
> 
> This error path leaks fd.

Patch sent.

Wei.

___
Xen-devel mailing list
Xen-devel@lists.xenproject.org
https://lists.xenproject.org/mailman/listinfo/xen-devel

Re: [Xen-devel] [PATCH] xenforeignmemory: work around bug in older privcmd

2018-08-28 Thread Andrew Cooper 
On 24/08/18 13:16, Paul Durrant wrote:
> diff --git a/tools/libs/foreignmemory/linux.c 
> b/tools/libs/foreignmemory/linux.c
> index 3686cf41e0..0368aa09f4 100644
> --- a/tools/libs/foreignmemory/linux.c
> +++ b/tools/libs/foreignmemory/linux.c
> @@ -53,6 +53,23 @@ int osdep_xenforeignmemory_open(xenforeignmemory_handle 
> *fmem)
>  return -1;
>  }
>  
> +/*
> + * Older versions of privcmd return -EINVAL for unimplemented ioctls
> + * so we need to probe for the errno to use rather than just using
> + * the conventional ENOTTY.
> + */
> +if ( ioctl(fd, IOCTL_PRIVCMD_UNIMPLEMENTED, NULL) >= 0 )
> +{
> +xtl_log(fmem->logger, XTL_ERROR, -1, "xenforeignmemory",
> +"privcmd ioctl should not be implemented");

This error path leaks fd.

~Andrew

> +return -1;
> +}
> +else
> +{
> +fmem->unimpl_errno = errno;
> +errno = 0;
> +}
> +
>  fmem->fd = fd;
>  return 0;
>  }
>


___
Xen-devel mailing list
Xen-devel@lists.xenproject.org
https://lists.xenproject.org/mailman/listinfo/xen-devel

Re: [Xen-devel] [PATCH] xenforeignmemory: work around bug in older privcmd

2018-08-27 Thread Wei Liu 
On Mon, Aug 27, 2018 at 03:04:55AM -0600, Jan Beulich wrote:
> >>> On 24.08.18 at 14:16,  wrote:
> > Versions of linux privcmd prior to commit dc9eab6fd94d ("return -ENOTTY
> > for unimplemented IOCTLs") will return -EINVAL rather than the conventional
> > -ENOTTY for unimplemented codes. This breaks the error path in
> > libxenforeignmemory resource mapping, which only translates ENOTTY into
> > EOPNOTSUPP to inform callers of the need to use an alternative (legacy)
> > mechanism.
> > 
> > This patch adds a new 'unimplemented' [1] ioctl code into the local
> > privcmd header which is then used to probe for the appropriate errno to
> > translate in the resource mapping error path
> > 
> > [1] this is a code that has, so far, never been used in any version of
> > privcmd and will be added to future versions of the header in the
> > linux source, to make sure it stays unimplemented.
> 
> Shouldn't this addition happen before the Xen tools side change goes
> in?

It doesn't really make any difference from a practical PoV. Applying
this patch right away helps unblocking various Linux branches.

Wei.

> 
> Jan
> 
> 

___
Xen-devel mailing list
Xen-devel@lists.xenproject.org
https://lists.xenproject.org/mailman/listinfo/xen-devel

Re: [Xen-devel] [PATCH] xenforeignmemory: work around bug in older privcmd

2018-08-27 Thread Jan Beulich 
>>> On 24.08.18 at 14:16,  wrote:
> Versions of linux privcmd prior to commit dc9eab6fd94d ("return -ENOTTY
> for unimplemented IOCTLs") will return -EINVAL rather than the conventional
> -ENOTTY for unimplemented codes. This breaks the error path in
> libxenforeignmemory resource mapping, which only translates ENOTTY into
> EOPNOTSUPP to inform callers of the need to use an alternative (legacy)
> mechanism.
> 
> This patch adds a new 'unimplemented' [1] ioctl code into the local
> privcmd header which is then used to probe for the appropriate errno to
> translate in the resource mapping error path
> 
> [1] this is a code that has, so far, never been used in any version of
> privcmd and will be added to future versions of the header in the
> linux source, to make sure it stays unimplemented.

Shouldn't this addition happen before the Xen tools side change goes
in?

Jan



___
Xen-devel mailing list
Xen-devel@lists.xenproject.org
https://lists.xenproject.org/mailman/listinfo/xen-devel

Re: [Xen-devel] [PATCH] xenforeignmemory: work around bug in older privcmd

2018-08-24 Thread Ian Jackson 
Paul Durrant writes ("[PATCH] xenforeignmemory: work around bug in older 
privcmd"):
> Versions of linux privcmd prior to commit dc9eab6fd94d ("return -ENOTTY
> for unimplemented IOCTLs") will return -EINVAL rather than the conventional
> -ENOTTY for unimplemented codes. This breaks the error path in
> libxenforeignmemory resource mapping, which only translates ENOTTY into
> EOPNOTSUPP to inform callers of the need to use an alternative (legacy)
> mechanism.
> 
> This patch adds a new 'unimplemented' [1] ioctl code into the local
> privcmd header which is then used to probe for the appropriate errno to
> translate in the resource mapping error path
> 
> [1] this is a code that has, so far, never been used in any version of
> privcmd and will be added to future versions of the header in the
> linux source, to make sure it stays unimplemented.

How irritating!  Thanks for fixing it.

Acked-by: Ian Jackson 

___
Xen-devel mailing list
Xen-devel@lists.xenproject.org
https://lists.xenproject.org/mailman/listinfo/xen-devel

Re: [Xen-devel] [PATCH] xenforeignmemory: work around bug in older privcmd

2018-08-24 Thread Wei Liu 
On Fri, Aug 24, 2018 at 01:16:26PM +0100, Paul Durrant wrote:
> Versions of linux privcmd prior to commit dc9eab6fd94d ("return -ENOTTY
> for unimplemented IOCTLs") will return -EINVAL rather than the conventional
> -ENOTTY for unimplemented codes. This breaks the error path in
> libxenforeignmemory resource mapping, which only translates ENOTTY into
> EOPNOTSUPP to inform callers of the need to use an alternative (legacy)
> mechanism.
> 
> This patch adds a new 'unimplemented' [1] ioctl code into the local
> privcmd header which is then used to probe for the appropriate errno to
> translate in the resource mapping error path
> 
> [1] this is a code that has, so far, never been used in any version of
> privcmd and will be added to future versions of the header in the
> linux source, to make sure it stays unimplemented.
> 
> Signed-off-by: Paul Durrant 
> ---

Acked-by: Wei Liu 

___
Xen-devel mailing list
Xen-devel@lists.xenproject.org
https://lists.xenproject.org/mailman/listinfo/xen-devel

[Xen-devel] [PATCH] xenforeignmemory: work around bug in older privcmd

2018-08-24 Thread Paul Durrant 
Versions of linux privcmd prior to commit dc9eab6fd94d ("return -ENOTTY
for unimplemented IOCTLs") will return -EINVAL rather than the conventional
-ENOTTY for unimplemented codes. This breaks the error path in
libxenforeignmemory resource mapping, which only translates ENOTTY into
EOPNOTSUPP to inform callers of the need to use an alternative (legacy)
mechanism.

This patch adds a new 'unimplemented' [1] ioctl code into the local
privcmd header which is then used to probe for the appropriate errno to
translate in the resource mapping error path

[1] this is a code that has, so far, never been used in any version of
privcmd and will be added to future versions of the header in the
linux source, to make sure it stays unimplemented.

Signed-off-by: Paul Durrant 
---
Cc: Ian Jackson 
Cc: Wei Liu 
---
 tools/include/xen-sys/Linux/privcmd.h |  2 ++
 tools/libs/foreignmemory/linux.c  | 19 ++-
 tools/libs/foreignmemory/private.h|  1 +
 3 files changed, 21 insertions(+), 1 deletion(-)

diff --git a/tools/include/xen-sys/Linux/privcmd.h 
b/tools/include/xen-sys/Linux/privcmd.h
index 9531b728f9..bc60e8fd55 100644
--- a/tools/include/xen-sys/Linux/privcmd.h
+++ b/tools/include/xen-sys/Linux/privcmd.h
@@ -114,5 +114,7 @@ typedef struct privcmd_mmap_resource {
_IOC(_IOC_NONE, 'P', 6, sizeof(domid_t))
 #define IOCTL_PRIVCMD_MMAP_RESOURCE\
_IOC(_IOC_NONE, 'P', 7, sizeof(privcmd_mmap_resource_t))
+#define IOCTL_PRIVCMD_UNIMPLEMENTED\
+   _IOC(_IOC_NONE, 'P', 0xFF, 0)
 
 #endif /* __LINUX_PUBLIC_PRIVCMD_H__ */
diff --git a/tools/libs/foreignmemory/linux.c b/tools/libs/foreignmemory/linux.c
index 3686cf41e0..0368aa09f4 100644
--- a/tools/libs/foreignmemory/linux.c
+++ b/tools/libs/foreignmemory/linux.c
@@ -53,6 +53,23 @@ int osdep_xenforeignmemory_open(xenforeignmemory_handle 
*fmem)
 return -1;
 }
 
+/*
+ * Older versions of privcmd return -EINVAL for unimplemented ioctls
+ * so we need to probe for the errno to use rather than just using
+ * the conventional ENOTTY.
+ */
+if ( ioctl(fd, IOCTL_PRIVCMD_UNIMPLEMENTED, NULL) >= 0 )
+{
+xtl_log(fmem->logger, XTL_ERROR, -1, "xenforeignmemory",
+"privcmd ioctl should not be implemented");
+return -1;
+}
+else
+{
+fmem->unimpl_errno = errno;
+errno = 0;
+}
+
 fmem->fd = fd;
 return 0;
 }
@@ -307,7 +324,7 @@ int osdep_xenforeignmemory_map_resource(
 {
 int saved_errno;
 
-if ( errno != ENOTTY && errno != EOPNOTSUPP )
+if ( errno != fmem->unimpl_errno && errno != EOPNOTSUPP )
 PERROR("ioctl failed");
 else
 errno = EOPNOTSUPP;
diff --git a/tools/libs/foreignmemory/private.h 
b/tools/libs/foreignmemory/private.h
index b06ce12583..8f1bf081ed 100644
--- a/tools/libs/foreignmemory/private.h
+++ b/tools/libs/foreignmemory/private.h
@@ -23,6 +23,7 @@ struct xenforeignmemory_handle {
 unsigned flags;
 int fd;
 Xentoolcore__Active_Handle tc_ah;
+int unimpl_errno;
 };
 
 int osdep_xenforeignmemory_open(xenforeignmemory_handle *fmem);
-- 
2.11.0


___
Xen-devel mailing list
Xen-devel@lists.xenproject.org
https://lists.xenproject.org/mailman/listinfo/xen-devel