[PATCH] powerpc: Initialize local variable fdt to NULL in elf64_load()

Dan Carpenter dan.carpenter at oracle.com
Fri Apr 16 18:09:41 AEST 2021


On Fri, Apr 16, 2021 at 09:00:12AM +0200, Christophe Leroy wrote:
> 
> 
> Le 16/04/2021 à 08:44, Daniel Axtens a écrit :
> > Hi Lakshmi,
> > 
> > > On 4/15/21 12:14 PM, Lakshmi Ramasubramanian wrote:
> > > 
> > > Sorry - missed copying device-tree and powerpc mailing lists.
> > > 
> > > > There are a few "goto out;" statements before the local variable "fdt"
> > > > is initialized through the call to of_kexec_alloc_and_setup_fdt() in
> > > > elf64_load(). This will result in an uninitialized "fdt" being passed
> > > > to kvfree() in this function if there is an error before the call to
> > > > of_kexec_alloc_and_setup_fdt().
> > > > 
> > > > Initialize the local variable "fdt" to NULL.
> > > > 
> > I'm a huge fan of initialising local variables! But I'm struggling to
> > find the code path that will lead to an uninit fdt being returned...
> > 
> > The out label reads in part:
> > 
> > 	/* Make kimage_file_post_load_cleanup free the fdt buffer for us. */
> > 	return ret ? ERR_PTR(ret) : fdt;
> > 
> > As far as I can tell, any time we get a non-zero ret, we're going to
> > return an error pointer rather than the uninitialised value...
> 
> I don't think GCC is smart enough to detect that.
> 

We disabled uninitialized variable checking for GCC.

But actually is something that has been on my mind recently.  Smatch is
supposed to parse this correctly but there is a bug that affects powerpc
and I don't know how to debug it.  The kbuild bot is doing cross
platform compiles but I don't have one set up on myself.  Could someone
with Smatch installed test something for me?

Or if you don't have Smatch installed then you should definitely install
it.  :P
https://www.spinics.net/lists/smatch/msg00568.html

Apply the patch from below and edit the path to point to the correct
directory.  Then run kchecker and email me the output?

~/path/to/smatch_scripts/kchecker arch/powerpc/kernel/hw_breakpoint.c

regads,
dan carpenter

diff --git a/arch/powerpc/kernel/hw_breakpoint.c b/arch/powerpc/kernel/hw_breakpoint.c
index 8fc7a14e4d71..f2dfba54e14d 100644
--- a/arch/powerpc/kernel/hw_breakpoint.c
+++ b/arch/powerpc/kernel/hw_breakpoint.c
@@ -167,13 +167,19 @@ static bool can_co_exist(struct breakpoint *b, struct perf_event *bp)
 	return !(alternate_infra_bp(b, bp) && bp_addr_range_overlap(b->bp, bp));
 }
 
+#include "/home/XXX/path/to/smatch/check_debug.h"
 static int task_bps_add(struct perf_event *bp)
 {
 	struct breakpoint *tmp;
 
 	tmp = alloc_breakpoint(bp);
-	if (IS_ERR(tmp))
+	__smatch_about(tmp);
+	__smatch_debug_on();
+	if (IS_ERR(tmp)) {
+		__smatch_debug_off();
+		__smatch_about(tmp);
 		return PTR_ERR(tmp);
+	}
 
 	list_add(&tmp->list, &task_bps);
 	return 0;


More information about the Linuxppc-dev mailing list