On Wed, Nov 18, 2015 at 1:07 PM, Shi, Yang <yang....@linaro.org> wrote: > On 11/18/2015 12:56 AM, Zi Shen Lim wrote: >> emit_a64_mov_i64(r3, size, ctx); >> - emit(A64_ADD_I(1, r4, fp, MAX_BPF_STACK), ctx); >> + emit(A64_SUB_I(1, r4, fp, STACK_SIZE), ctx); > > > Should not it sub MAX_BPF_STACK?
No, if it's at (BPF_FP - MAX_BPF_STACK), we'll be writing into the BPF stack area, which should only be used by the BPF program. > If you sub STACK_SIZE here, the buffer pointer will point to bottom of the > reserved area. Yes, that's the idea. The buffer is allocated in here. Right now we're using this "reserved" space for this buffer only. > > You stack layout change also shows this: > > + * +-----+ <= (BPF_FP - MAX_BPF_STACK) > + * |RSVD | JIT scratchpad > + * current A64_SP => +-----+ <= (BPF_FP - STACK_SIZE) Yes, this diagram reflects the code and intention. Thanks for reviewing, we definitely need more of these :) -- To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to majord...@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html Please read the FAQ at http://www.tux.org/lkml/
