[PATCH] x86/intel_rdt: fix rdt_mount error handling

From: Arnd Bergmann
Date: Tue Nov 08 2016 - 08:44:40 EST


The newly introduced rdt_mount function returns an unintialized
pointer if rdtgroup_create_info_dir() fails:

arch/x86/kernel/cpu/intel_rdt_rdtgroup.c: In function ârdt_mountâ:
arch/x86/kernel/cpu/intel_rdt_rdtgroup.c:710:9: error: âdentryâ may be used uninitialized in this function [-Werror=maybe-uninitialized]

Clearly the intention was to propagate the error code here as we
do in the other failure cases.

Fixes: 5ff193fbde20 ("x86/intel_rdt: Add basic resctrl filesystem support")
Signed-off-by: Arnd Bergmann <arnd@xxxxxxxx>
---
arch/x86/kernel/cpu/intel_rdt_rdtgroup.c | 4 +++-
1 file changed, 3 insertions(+), 1 deletion(-)

diff --git a/arch/x86/kernel/cpu/intel_rdt_rdtgroup.c b/arch/x86/kernel/cpu/intel_rdt_rdtgroup.c
index a90ad22b9823..e66c7a58505e 100644
--- a/arch/x86/kernel/cpu/intel_rdt_rdtgroup.c
+++ b/arch/x86/kernel/cpu/intel_rdt_rdtgroup.c
@@ -691,8 +691,10 @@ static struct dentry *rdt_mount(struct file_system_type *fs_type,
closid_init();

ret = rdtgroup_create_info_dir(rdtgroup_default.kn);
- if (ret)
+ if (ret) {
+ dentry = ERR_PTR(ret);
goto out_cdp;
+ }

dentry = kernfs_mount(fs_type, flags, rdt_root,
RDTGROUP_SUPER_MAGIC, NULL);
--
2.9.0