Messages in this thread | | | Date | Tue, 15 Sep 2020 10:26:17 -0300 | From | "Daniel W. S. Almeida" <> | Subject | Re: [v10 3/4] media: vidtv: add a bridge driver |
| |
Hi Geert,
Thanks for bringing this to my attention.
>> + u32 nbytes = 0; /* the number of bytes written by this function */ >> + >> + u64 nbytes_expected; /* the number of bytes we should have written */ >> + u64 nbytes_streamed; /* the number of bytes we actually wrote */ >> + u32 num_null_pkts; /* number of null packets to bridge the gap */ >> + >> + u64 elapsed_time_msecs = jiffies_to_usecs(m->timing.current_jiffies - >> + m->timing.past_jiffies); >> + >> + elapsed_time_msecs = min(elapsed_time_msecs, >> (u64)VIDTV_MAX_SLEEP_USECS / 1000); >> + nbytes_expected = div64_u64(m->mux_rate_kbytes_sec * 1000, MSEC_PER_SEC); > > Seriously?!? > > You multiply by 1000 first, followed by a division by 1000 using an > expensive 64-by-64 division?
This entire function is broken and needs a do-over :)
> using an expensive 64-by-64 division?
I am new to kernel development. I wasn't even aware that this was expensive, to be honest.
>> + if (nbytes_streamed < nbytes_expected) { >> + /* can't write half a packet: roundup to a 188 multiple */ >> + nbytes_expected = roundup(nbytes_expected - nbytes_streamed, TS_PACKET_LEN); > > drivers/media/test-drivers/vidtv/vidtv_mux.o: In function `vidtv_mux_tick': > vidtv_mux.c:(.text+0x788): undefined reference to `__udivdi3' > > This is a 64-by-32 division, hence it should use a helper from > <linux/math64.h>. > > However, I'm wondering if "nbytes_expected - nbytes_streamed" is > guaranteed to be a "small" number, hence a 32-by-32 division would be > sufficient?
I think so.
I will send a patch to address the things you pointed out in this email.
-- thanks -- Daniel
| |