From: Cyril Hrubis <chrubis@suse.cz>
To: ltp@lists.linux.it
Subject: [LTP] [PATCH v4 1/5] libs/libltpnewipc/libnewipc.c: Add msg_do_reader/msg_do_writer function
Date: Wed, 21 Jul 2021 17:29:56 +0200 [thread overview]
Message-ID: <YPg9dN2GhnncVnQQ@yuki> (raw)
In-Reply-To: <1615550541-21714-1-git-send-email-xuyang2018.jy@cn.fujitsu.com>
Hi!
> +void msg_do_reader(long key, int tid, long type, int child, int nreps)
> +{
> + int i, size;
> + int id;
> + struct mbuffer buffer;
> +
> + id = SAFE_MSGGET(key, 0);
> + if (id != tid) {
> + tst_brk(TFAIL,
> + "Message queue mismatch in the reader of child group %d for message queue id %d ",
> + child, id);
> + }
Also looking at how these functions are used, we fork a child that calls
this function then wait it and ignore the exit value it does not make
sense to use tst_brk() here at all.
The tst_brk() is supposed to exit the whole test, including all
suprocesses but that does not work if we throw away the child return
value. This piece of code looks like we exit the whole test here, which
isn't simply true.
So either we should do tst_res(TFAIL, ...) followed by an exit(0) here,
or we should handle the return value in the parent. I.e. stop the test
if one of the children reported a failure.
> + for (i = 0; i < nreps; i++) {
> + memset(&buffer, 0, sizeof(buffer));
> +
> + size = SAFE_MSGRCV(id, &buffer, 100, type, 0);
> + if (buffer.type != type) {
> + tst_brk(TFAIL,
> + "Type mismatch in child %d, read #%d, for message got %ld, exected %ld",
> + child, (i + 1), buffer.type, type);
> + }
> + if (buffer.data.len + 1 != size) {
> + tst_brk(TFAIL,
> + "Size mismatch in child %d, read #%d, for message got %d, expected %d",
> + child, (i + 1), buffer.data.len + 1, size);
> + }
> + if (verify(buffer.data.pbytes, (key % 255), size - 1, child)) {
> + tst_brk(TFAIL,
> + "Verify failed in child %d read # = %d, key = %lx",
> + child, (i + 1), key);
> + }
> + key++;
> + }
It would also make sense to do exit(0); here so that we do not have to
repeat that in each test.
> +}
> +
> +void msg_do_writer(long key, int tid, long type, int child, int nreps)
> +{
> + int i, size;
> + int id;
> + struct mbuffer buffer;
> +
> + id = SAFE_MSGGET(key, 0);
> + if (id != tid) {
> + tst_brk(TFAIL,
> + "Message queue mismatch in the writer of child group %d for message queue id %d",
> + child, id);
> + }
> +
> + for (i = 0; i < nreps; i++) {
> + memset(&buffer, 0, sizeof(buffer));
> +
> + do {
> + size = (lrand48() % 99);
> + } while (size == 0);
> + memset(buffer.data.pbytes, (key % 255), size);
> + buffer.data.len = size;
> + buffer.type = type;
> + SAFE_MSGSND(id, &buffer, size + 1, 0);
> + key++;
> + }
> +}
And same here.
--
Cyril Hrubis
chrubis@suse.cz
next prev parent reply other threads:[~2021-07-21 15:29 UTC|newest]
Thread overview: 26+ messages / expand[flat|nested] mbox.gz Atom feed top
2020-06-18 9:41 [LTP] [PATCH v2 1/2] libs/libltpnewipc: Add libnewmsgctl.c Yang Xu
2020-06-18 9:41 ` [LTP] [PATCH v2 2/2] syscalls/msgstress: scale number of processes accodingly to available RAM Yang Xu
2020-07-22 15:45 ` Cyril Hrubis
2020-07-23 6:34 ` Yang Xu
2020-10-21 12:57 ` [LTP] [PATCH v3 1/4] syscalls/msgstress: Add common file for msgstress Yang Xu
2020-10-21 12:57 ` [LTP] [PATCH v3 2/4] syscalls/msgstress*: cleanup and convert into new api Yang Xu
2020-11-11 16:31 ` Cyril Hrubis
2021-03-12 12:02 ` [LTP] [PATCH v4 1/5] libs/libltpnewipc/libnewipc.c: Add msg_do_reader/msg_do_writer function Yang Xu
2021-03-12 12:02 ` [LTP] [PATCH v4 2/5] syscalls/msgstress01: Convert into new api and merge msgstress03 into it Yang Xu
2021-07-21 15:07 ` Cyril Hrubis
2021-03-12 12:02 ` [LTP] [PATCH v4 3/5] syscalls/msgstress02: Convert into new api Yang Xu
2021-07-21 15:33 ` Cyril Hrubis
2021-03-12 12:02 ` [LTP] [PATCH v4 4/5] sycalls/msgstress04: " Yang Xu
2021-07-21 15:35 ` Cyril Hrubis
2021-03-12 12:02 ` [LTP] [PATCH v4 5/5] libs/libltpipc: Remove useless function and c file Yang Xu
2021-07-21 14:27 ` [LTP] [PATCH v4 1/5] libs/libltpnewipc/libnewipc.c: Add msg_do_reader/msg_do_writer function Cyril Hrubis
2021-07-21 15:29 ` Cyril Hrubis [this message]
2020-10-21 12:57 ` [LTP] [PATCH v3 3/4] libmsgctl: Remove it Yang Xu
2020-10-21 12:57 ` [LTP] [PATCH v3 4/4] lipipc: Remove useless get_max_msgqueues api Yang Xu
2020-11-11 15:44 ` [LTP] [PATCH v3 1/4] syscalls/msgstress: Add common file for msgstress Cyril Hrubis
2020-11-13 6:14 ` Yang Xu
2020-11-13 15:26 ` Cyril Hrubis
2020-06-24 5:04 ` [LTP] [PATCH v2 1/2] libs/libltpnewipc: Add libnewmsgctl.c Yang Xu
2020-07-16 1:07 ` Yang Xu
2020-07-22 15:01 ` Cyril Hrubis
2020-07-23 6:29 ` Yang Xu
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=YPg9dN2GhnncVnQQ@yuki \
--to=chrubis@suse.cz \
--cc=ltp@lists.linux.it \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox