On Tue, Aug 26, 2025 at 3:08 PM Tiezhu Yang <yangtiezhu@xxxxxxxxxxx> wrote:
>
> On 2025/8/19 下午7:19, Huacai Chen wrote:
> > In __arch_prepare_bpf_trampoline(), retval_off is meaningful only when
> > save_ret is not 0, so the current logic is correct. But it may cause a
> > build warning:
> >
> > arch/loongarch/net/bpf_jit.c:1547 __arch_prepare_bpf_trampoline() error: uninitialized symbol 'retval_off'.
> >
> > So initialize retval_off unconditionally to fix it.
> >
> > Fixes: f9b6b41f0cf3 ("LoongArch: BPF: Add basic bpf trampoline support")
> > Closes: https://lore.kernel.org/r/202508191020.PBBh07cK-lkp@xxxxxxxxx/
> > Reported-by: kernel test robot <lkp@xxxxxxxxx>
> > Reported-by: Dan Carpenter <dan.carpenter@xxxxxxxxxx>
> > Signed-off-by: Huacai Chen <chenhuacai@xxxxxxxxxxx>
> > ---
> > arch/loongarch/net/bpf_jit.c | 9 ++++-----
> > 1 file changed, 4 insertions(+), 5 deletions(-)
> >
> > diff --git a/arch/loongarch/net/bpf_jit.c b/arch/loongarch/net/bpf_jit.c
> > index abfdb6bb5c38..a73f6ea4ed4a 100644
> > --- a/arch/loongarch/net/bpf_jit.c
> > +++ b/arch/loongarch/net/bpf_jit.c
> > @@ -1504,11 +1504,10 @@ static int __arch_prepare_bpf_trampoline(struct jit_ctx *ctx, struct bpf_tramp_i
> > stack_size += 16;
> >
> > save_ret = flags & (BPF_TRAMP_F_CALL_ORIG | BPF_TRAMP_F_RET_FENTRY_RET);
> > - if (save_ret) {
> > - /* Save BPF R0 and A0 */
> > - stack_size += 16;
> > - retval_off = stack_size;
> > - }
> > + if (save_ret)
> > + stack_size += 16; /* Save BPF R0 and A0 */
> > +
> > + retval_off = stack_size;
>
> Just init retval_off as 0 at the beginning of this function?
> What is the difference? which is better?
This patch is more like ARM64 does.
In addition, if save_ret is true, "init retval_off as 0" causes
retval_off to be evaluated twice, while this patch is only evaluated
once. But surely this is not a big deal.
Huacai
>
> Thanks,
> Tiezhu
>
