Re: [PATCH v2] tracing: Introduce pipe_cpumask to avoid race on trace_pipes

[Google]

On Sat, 19 Aug 2023 10:42:57 +0900
Masami Hiramatsu (Google) <mhiramat@xxxxxxxxxx> wrote:

> On Fri, 18 Aug 2023 11:53:22 -0400
> Steven Rostedt <rostedt@xxxxxxxxxxx> wrote:
> 
> > On Fri, 18 Aug 2023 23:23:01 +0900
> > Masami Hiramatsu (Google) <mhiramat@xxxxxxxxxx> wrote:
> > 
> > > It uses trace_pipe_raw. I guess if splice(from trace_pipe_raw to virtio-serial)
> > > returns -1 and errno == EAGAIN, the trace data will be lost?
> > 
> > It shouldn't. If it does, then there's likely a bug. The code will block
> > and if an interrupt comes in it will return immediately without reading
> > from the buffer.
> > 
> > https://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git/tree/kernel/trace/trace.c#n8262
> > 
> > I don't see where it would return -EINTR and consume data, but I may be
> > missing something.
> 
> Hmm, I suspect the case if the spilice_to_pipe() returns -EAGAIN.
> 
> https://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git/tree/kernel/trace/trace.c#n8491
> 
> It seems not handling such case.
> 
> Anyway, I also think something wrong in virtio-serial (or misusing?), since
> it can not read anything from the host sometimes. I just setup the virtio-trace
> with below patch (ignore EAGAIN).
> 
> 
> From 92242480285448360c9390a743ea7b3751bb3e61 Mon Sep 17 00:00:00 2001
> From: "Masami Hiramatsu (Google)" <mhiramat@xxxxxxxxxx>
> Date: Thu, 17 Aug 2023 14:08:40 +0900
> Subject: [PATCH 1/3] tools/virtio-trace: Ignore EAGAIN error on splice()
> 
> splice() can return EAGAIN error instead of returning 0 size read.
> In that case, wait a while and try to call splice() again.
> 
> Signed-off-by: Masami Hiramatsu (Google) <mhiramat@xxxxxxxxxx>
> ---
>  tools/virtio/virtio-trace/trace-agent-rw.c | 7 ++++---
>  1 file changed, 4 insertions(+), 3 deletions(-)
> 
> diff --git a/tools/virtio/virtio-trace/trace-agent-rw.c b/tools/virtio/virtio-trace/trace-agent-rw.c
> index ddfe7875eb16..e8a4c4f0c499 100644
> --- a/tools/virtio/virtio-trace/trace-agent-rw.c
> +++ b/tools/virtio/virtio-trace/trace-agent-rw.c
> @@ -8,6 +8,7 @@
>   */
>  
>  #define _GNU_SOURCE
> +#include <errno.h>
>  #include <fcntl.h>
>  #include <stdio.h>
>  #include <stdlib.h>
> @@ -127,10 +128,10 @@ static void *rw_thread_main(void *thread_info)
>  		rlen = splice(ts->in_fd, NULL, ts->read_pipe, NULL,
>  				ts->pipe_size, SPLICE_F_MOVE | SPLICE_F_MORE);
>  
> -		if (rlen < 0) {
> -			pr_err("Splice_read in rw-thread(%d)\n", ts->cpu_num);
> +		if (rlen < 0 && errno != EAGAIN) {
> +			pr_err("Splice_read error (%d) in rw-thread(%d)\n", errno, ts->cpu_num);
>  			goto error;
> -		} else if (rlen == 0) {
> +		} else if (rlen == 0 || errno == EAGAIN) {

Ah, this caused a drop. errno can be EAGAIN even if rlen > 0.
I've fixed this and that works.
BTW, I think this virtio-trace would be better to move under
tools/tracing because it is a tracing tool.

Thanks,

>  			/*
>  			 * If trace data do not exist or are unreadable not
>  			 * for exceeding the page size, splice_read returns
> -- 
> 2.34.1
> 
> 
> -- 
> Masami Hiramatsu (Google) <mhiramat@xxxxxxxxxx>


-- 
Masami Hiramatsu (Google) <mhiramat@xxxxxxxxxx>