On Fri, Sep 23, 2016 at 4:38 PM, Paul Burton <paul.burton@xxxxxxxxxx> wrote: > When allocating a struct alias_prop, of_alias_scan() only requested that > it be aligned on a 4 byte boundary. The struct contains pointers which > leads to us attempting 64 bit writes on 64 bit systems, and if the CPU > doesn't support unaligned memory accesses then this causes problems - > for example on some MIPS64r2 CPUs including the "mips64r2-generic" QEMU > emulated CPU it will trigger an address error exception. > > Fix this by requesting alignment for the struct alias_prop allocation > matching that which the compiler expects, using the __alignof__ keyword. > > Signed-off-by: Paul Burton <paul.burton@xxxxxxxxxx> Looks right to me. Reviewed-by: Grant Likely <grant.likely@xxxxxxxxxxxx> > Cc: devicetree@xxxxxxxxxxxxxxx > Cc: Rob Herring <robh+dt@xxxxxxxxxx> > Cc: Frank Rowand <frowand.list@xxxxxxxxx> > Cc: linux-kernel@xxxxxxxxxxxxxxx > Cc: linux-mips@xxxxxxxxxxxxxx > Cc: Ralf Baechle <ralf@xxxxxxxxxxxxxx> > --- > drivers/of/base.c | 2 +- > 1 file changed, 1 insertion(+), 1 deletion(-) > > diff --git a/drivers/of/base.c b/drivers/of/base.c > index 3ce6953..448aa40 100644 > --- a/drivers/of/base.c > +++ b/drivers/of/base.c > @@ -2042,7 +2042,7 @@ void of_alias_scan(void * (*dt_alloc)(u64 size, u64 align)) > continue; > > /* Allocate an alias_prop with enough space for the stem */ > - ap = dt_alloc(sizeof(*ap) + len + 1, 4); > + ap = dt_alloc(sizeof(*ap) + len + 1, __alignof__(*ap)); > if (!ap) > continue; > memset(ap, 0, sizeof(*ap) + len + 1); > -- > 2.10.0 > -- To unsubscribe from this list: send the line "unsubscribe devicetree" in the body of a message to majordomo@xxxxxxxxxxxxxxx More majordomo info at http://vger.kernel.org/majordomo-info.html
