From: "Knop, Ryszard" <ryszard.knop@intel.com>
To: "igt-dev@lists.freedesktop.org" <igt-dev@lists.freedesktop.org>,
"kamil.konieczny@linux.intel.com"
<kamil.konieczny@linux.intel.com>
Subject: Re: [i-g-t,v1,2/2] runner: Create new option for incremental results
Date: Tue, 21 Jul 2026 13:20:02 +0000 [thread overview]
Message-ID: <ab3f212acab766cd62f3ab5d84489cc387653563.camel@intel.com> (raw)
In-Reply-To: <20260720161526.64709-3-kamil.konieczny@linux.intel.com>
On Mon, 2026-07-20 at 18:15 +0200, Kamil Konieczny wrote:
> Create new option --inc-results which will make results.json
> after each test was run, not only after last one.
>
> Please note that this could be somewhat misleading when a kernel
> panic happens during an execution, as a real test causing this
> will not be there. In such cases one still needs to run
> igt_results to add that last one to results.
>
> Signed-off-by: Kamil Konieczny <kamil.konieczny@linux.intel.com>
> ---
> runner/executor.c | 19 +++++++++++++------
> runner/settings.c | 8 ++++++++
> runner/settings.h | 1 +
> 3 files changed, 22 insertions(+), 6 deletions(-)
>
> diff --git a/runner/executor.c b/runner/executor.c
> index b204fdffc..7c923bed3 100644
> --- a/runner/executor.c
> +++ b/runner/executor.c
> @@ -2045,6 +2045,11 @@ static bool clear_old_results(char *path)
> return false;
> }
>
> + if (remove_file(dirfd, "results.old")) {
> + errf("Error clearing results.old : %m\n");
> + return false;
> + }
> +
> for (i = 0; true; i++) {
> struct stat st;
>
> @@ -2484,6 +2489,9 @@ static bool generate_results_with_endtime(int resdirfd, struct settings *setting
> double end_time;
> bool ret;
>
> + if (!settings->inc_results)
> + return true;
> +
> remove_file(resdirfd, "results.bak");
> renameat(resdirfd, "results.json", resdirfd, "results.bak");
>
> @@ -2758,12 +2766,11 @@ bool execute(struct execute_state *state,
> return false;
>
> return execute(state, settings, job_list);
> - }
> -
> - /* it should check option, also measure time taken */
> - if (!generate_results_with_endtime(resdirfd, settings)) {
> - status = false;
> - break;
> + } else {
> + if (!generate_results_with_endtime(resdirfd, settings)) {
> + status = false;
> + break;
> + }
> }
> }
>
> diff --git a/runner/settings.c b/runner/settings.c
> index 3ccf71870..c52fed92b 100644
> --- a/runner/settings.c
> +++ b/runner/settings.c
> @@ -23,6 +23,7 @@ enum {
> OPT_DISK_USAGE_LIMIT,
> OPT_TEST_LIST,
> OPT_IGNORE_MISSING,
> + OPT_INC_RESULTS,
> OPT_PIGLIT_DMESG,
> OPT_DMESG_WARN_LEVEL,
> OPT_OVERALL_TIMEOUT,
> @@ -309,6 +310,7 @@ static const char *usage_str =
> " Set an environment variable for the test process.\n"
> " If only the key is provided, the current value is read\n"
> " from the runner's environment (and saved for resumes).\n"
> + " --inc-results Save results incrementally after each test was run\n"
Can this be named something like "--save-results-after-each-test" or
something like that? It's not really "incremental", the entire
results.json is getting generated from scratch on each run, and a
longer name is a bit more descriptive.
> " -L, --list-all List all matching subtests instead of running\n"
> " --collect-code-cov Enables gcov-based collect of code coverage for tests.\n"
> " Requires --collect-script FILENAME\n"
> @@ -693,6 +695,7 @@ bool parse_options(int argc, char **argv,
> {"abort-on-monitored-error", optional_argument, NULL, OPT_ABORT_ON_ERROR},
> {"disk-usage-limit", required_argument, NULL, OPT_DISK_USAGE_LIMIT},
> {"facts", no_argument, NULL, OPT_FACTS},
> + {"inc-results", no_argument, NULL, OPT_INC_RESULTS},
> {"kmemleak", optional_argument, NULL, OPT_KMEMLEAK},
> {"sync", no_argument, NULL, OPT_SYNC},
> {"log-level", required_argument, NULL, OPT_LOG_LEVEL},
> @@ -768,6 +771,9 @@ bool parse_options(int argc, char **argv,
> case OPT_FACTS:
> settings->facts = true;
> break;
> + case OPT_INC_RESULTS:
> + settings->inc_results = true;
> + break;
> case OPT_KMEMLEAK:
> /* The default is once */
> settings->kmemleak = true;
> @@ -1235,6 +1241,7 @@ bool serialize_settings(struct settings *settings)
> SERIALIZE_INT(f, settings, dry_run);
> SERIALIZE_INT(f, settings, allow_non_root);
> SERIALIZE_INT(f, settings, facts);
> + SERIALIZE_INT(f, settings, inc_results);
> SERIALIZE_INT(f, settings, kmemleak);
> SERIALIZE_INT(f, settings, kmemleak_each);
> SERIALIZE_INT(f, settings, sync);
> @@ -1353,6 +1360,7 @@ bool read_settings_from_file(struct settings *settings, FILE *f)
> PARSE_INT(settings, name, val, dry_run);
> PARSE_INT(settings, name, val, allow_non_root);
> PARSE_INT(settings, name, val, facts);
> + PARSE_INT(settings, name, val, inc_results);
> PARSE_INT(settings, name, val, kmemleak);
> PARSE_INT(settings, name, val, kmemleak_each);
> PARSE_INT(settings, name, val, sync);
> diff --git a/runner/settings.h b/runner/settings.h
> index 6c58c3282..10f95f286 100644
> --- a/runner/settings.h
> +++ b/runner/settings.h
> @@ -62,6 +62,7 @@ struct settings {
> struct igt_list_head env_vars;
> struct igt_vec hook_strs;
> bool facts;
> + bool inc_results;
> bool kmemleak;
> bool kmemleak_each;
> bool sync;
Thanks, Ryszard
next prev parent reply other threads:[~2026-07-21 13:20 UTC|newest]
Thread overview: 6+ messages / expand[flat|nested] mbox.gz Atom feed top
2026-07-20 16:15 [PATCH i-g-t RFC v1 0/2] runner: create results incrementally Kamil Konieczny
2026-07-20 16:15 ` [PATCH i-g-t v1 1/2] runner: Create results after each tests end Kamil Konieczny
2026-07-21 13:17 ` [i-g-t,v1,1/2] " Knop, Ryszard
2026-07-20 16:15 ` [PATCH i-g-t v1 2/2] runner: Create new option for incremental results Kamil Konieczny
2026-07-21 13:20 ` Knop, Ryszard [this message]
2026-07-21 22:20 ` [PATCH i-g-t RFC v1 0/2] runner: create results incrementally Gustavo Sousa
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=ab3f212acab766cd62f3ab5d84489cc387653563.camel@intel.com \
--to=ryszard.knop@intel.com \
--cc=igt-dev@lists.freedesktop.org \
--cc=kamil.konieczny@linux.intel.com \
/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