From: Steven Rostedt <rostedt@goodmis.org>
To: Pengpeng Hou <pengpeng@iscas.ac.cn>
Cc: mhiramat@kernel.org, mathieu.desnoyers@efficios.com,
tom.zanussi@linux.intel.com, linux-kernel@vger.kernel.org,
linux-trace-kernel@vger.kernel.org
Subject: Re: [PATCH 2/2] tracing/hist: allocate synthetic-field command buffers to fit
Date: Sun, 29 Mar 2026 14:49:39 -0400 [thread overview]
Message-ID: <20260329144939.55cdf935@gandalf.local.home> (raw)
In-Reply-To: <20260329030950.32503-2-pengpeng@iscas.ac.cn>
On Sun, 29 Mar 2026 11:09:50 +0800
Pengpeng Hou <pengpeng@iscas.ac.cn> wrote:
> The synthetic field helpers currently build temporary names and trigger
> commands in fixed MAX_FILTER_STR_VAL buffers with strcpy() and strcat().
> Long field names, key lists, or saved filters can therefore overrun
> those staging buffers while constructing the synthetic histogram
> command.
>
> Allocate the synthetic name and command buffers to the exact size
> required by the current histogram instead of relying on fixed-size
> scratch storage.
No, the names should never be greater than the max value defined. If they
are, then it should error out.
>
> Fixes: 02205a6752f2 ("tracing: Add support for 'field variables'")
> Signed-off-by: Pengpeng Hou <pengpeng@iscas.ac.cn>
> ---
> kernel/trace/trace_events_hist.c | 46 +++++++++++++++++++-------------
> 1 file changed, 28 insertions(+), 18 deletions(-)
>
> diff --git a/kernel/trace/trace_events_hist.c b/kernel/trace/trace_events_hist.c
> index 4a27da628a71..1883bd6d9b95 100644
> --- a/kernel/trace/trace_events_hist.c
> +++ b/kernel/trace/trace_events_hist.c
> @@ -2964,13 +2964,10 @@ find_synthetic_field_var(struct hist_trigger_data *target_hist_data,
> struct hist_field *event_var;
> char *synthetic_name;
>
Should be preceded with:
if (strlen("synthetic_") + strlen(field_name) >= MAX_FILTER_STR_VAL)
return ERR_PTR(-E2BIG);
> - synthetic_name = kzalloc(MAX_FILTER_STR_VAL, GFP_KERNEL);
> + synthetic_name = kasprintf(GFP_KERNEL, "synthetic_%s", field_name);
> if (!synthetic_name)
> return ERR_PTR(-ENOMEM);
>
> - strcpy(synthetic_name, "synthetic_");
> - strcat(synthetic_name, field_name);
> -
> event_var = find_event_var(target_hist_data, system, event_name, synthetic_name);
>
> kfree(synthetic_name);
> @@ -3016,6 +3013,8 @@ create_field_var_hist(struct hist_trigger_data *target_hist_data,
> struct hist_field *event_var;
> char *saved_filter;
> char *cmd;
> + size_t cmdlen;
> + size_t off;
> int ret;
>
> if (target_hist_data->n_field_var_hists >= SYNTH_FIELDS_MAX) {
> @@ -3053,35 +3052,46 @@ create_field_var_hist(struct hist_trigger_data *target_hist_data,
> if (!var_hist)
> return ERR_PTR(-ENOMEM);
>
> - cmd = kzalloc(MAX_FILTER_STR_VAL, GFP_KERNEL);
> + saved_filter = find_trigger_filter(hist_data, file);
> +
> + cmdlen = strlen("keys=") + strlen(":synthetic_") +
> + strlen(field_name) + strlen("=") + strlen(field_name) + 1;
> + first = true;
> + for_each_hist_key_field(i, hist_data) {
> + key_field = hist_data->fields[i];
> + if (!first)
> + cmdlen++;
> + cmdlen += strlen(key_field->field->name);
> + first = false;
> + }
> +
> + if (saved_filter)
> + cmdlen += strlen(" if ") + strlen(saved_filter);
> +
Length should be checked. There shouldn't be huge strings for filters.
Perhaps in the future we may make them bigger but for now, 256 bytes should
be the limit.
-- Steve
> + cmd = kzalloc(cmdlen, GFP_KERNEL);
> if (!cmd) {
> kfree(var_hist);
> return ERR_PTR(-ENOMEM);
> }
>
> /* Use the same keys as the compatible histogram */
> - strcat(cmd, "keys=");
> + off = scnprintf(cmd, cmdlen, "keys=");
> + first = true;
>
> for_each_hist_key_field(i, hist_data) {
> key_field = hist_data->fields[i];
> - if (!first)
> - strcat(cmd, ",");
> - strcat(cmd, key_field->field->name);
> + off += scnprintf(cmd + off, cmdlen - off, "%s%s",
> + first ? "" : ",",
> key_field->field->name); first = false;
> }
>
> /* Create the synthetic field variable specification */
> - strcat(cmd, ":synthetic_");
> - strcat(cmd, field_name);
> - strcat(cmd, "=");
> - strcat(cmd, field_name);
> + off += scnprintf(cmd + off, cmdlen - off, ":synthetic_%s=%s",
> + field_name, field_name);
>
> /* Use the same filter as the compatible histogram */
> - saved_filter = find_trigger_filter(hist_data, file);
> - if (saved_filter) {
> - strcat(cmd, " if ");
> - strcat(cmd, saved_filter);
> - }
> + if (saved_filter)
> + scnprintf(cmd + off, cmdlen - off, " if %s",
> saved_filter);
> var_hist->cmd = kstrdup(cmd, GFP_KERNEL);
> if (!var_hist->cmd) {
next prev parent reply other threads:[~2026-03-29 18:48 UTC|newest]
Thread overview: 10+ messages / expand[flat|nested] mbox.gz Atom feed top
2026-03-29 3:09 [PATCH 2/2] tracing/hist: allocate synthetic-field command buffers to fit Pengpeng Hou
2026-03-29 18:49 ` Steven Rostedt [this message]
2026-03-30 2:46 ` [PATCH v2 2/2] tracing/hist: reject synthetic-field strings that exceed MAX_FILTER_STR_VAL Pengpeng Hou
2026-04-01 11:22 ` Pengpeng Hou
2026-04-08 21:31 ` Steven Rostedt
2026-04-09 2:19 ` [PATCH v3] tracing/hist: bound synthetic-field strings with seq_buf Pengpeng Hou
2026-04-14 8:58 ` Steven Rostedt
2026-04-17 3:06 ` Pengpeng Hou
2026-04-17 12:20 ` [PATCH v4] tracing: Bound " Pengpeng Hou
2026-04-17 16:16 ` Steven Rostedt
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=20260329144939.55cdf935@gandalf.local.home \
--to=rostedt@goodmis.org \
--cc=linux-kernel@vger.kernel.org \
--cc=linux-trace-kernel@vger.kernel.org \
--cc=mathieu.desnoyers@efficios.com \
--cc=mhiramat@kernel.org \
--cc=pengpeng@iscas.ac.cn \
--cc=tom.zanussi@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