From: Hao Ge <hao.ge@linux.dev>
To: Abhishek Bapat <abhishekbapat@google.com>,
Suren Baghdasaryan <surenb@google.com>,
Andrew Morton <akpm@linux-foundation.org>,
Kent Overstreet <kent.overstreet@linux.dev>
Cc: Shuah Khan <skhan@linuxfoundation.org>,
Jonathan Corbet <corbet@lwn.net>,
linux-doc@vger.kernel.org, linux-kernel@vger.kernel.org,
linux-mm@kvack.org, Sourav Panda <souravpanda@google.com>
Subject: Re: [PATCH v2 3/6] alloc_tag: add size-based filtering to ioctl
Date: Tue, 26 May 2026 11:11:33 +0800 [thread overview]
Message-ID: <de0f2984-44ea-4098-9d19-c63ee035cdaf@linux.dev> (raw)
In-Reply-To: <c4b425d1f9192caca3cad830f322aa048ed26d45.1779471082.git.abhishekbapat@google.com>
Hi Abhishek
On 2026/5/23 01:45, Abhishek Bapat wrote:
> Extend the allocinfo filtering mechanism to allow users to filter tags
> based on the total number of bytes allocated [min_size, max_size]. The
> size range is inclusive.
>
> Filtering by size involves retrieving allocinfo per-CPU counters, which
> is an expensive operation. Hence, the performance of size-based
> filtering will be worse than other filters.
>
> Signed-off-by: Abhishek Bapat <abhishekbapat@google.com>
> ---
> include/uapi/linux/alloc_tag.h | 8 +++-
> lib/alloc_tag.c | 72 ++++++++++++++++++++++++++++------
> 2 files changed, 68 insertions(+), 12 deletions(-)
>
> diff --git a/include/uapi/linux/alloc_tag.h b/include/uapi/linux/alloc_tag.h
> index 0cc9db5298c6..45f158bee0a6 100644
> --- a/include/uapi/linux/alloc_tag.h
> +++ b/include/uapi/linux/alloc_tag.h
> @@ -39,13 +39,17 @@ enum {
> ALLOCINFO_FILTER_FUNCTION,
> ALLOCINFO_FILTER_FILENAME,
> ALLOCINFO_FILTER_LINENO,
> - __ALLOCINFO_FILTER_LAST = ALLOCINFO_FILTER_LINENO
> + ALLOCINFO_FILTER_MIN_SIZE,
> + ALLOCINFO_FILTER_MAX_SIZE,
> + __ALLOCINFO_FILTER_LAST = ALLOCINFO_FILTER_MAX_SIZE
> };
>
> #define ALLOCINFO_FILTER_MASK_MODNAME (1 << ALLOCINFO_FILTER_MODNAME)
> #define ALLOCINFO_FILTER_MASK_FUNCTION (1 << ALLOCINFO_FILTER_FUNCTION)
> #define ALLOCINFO_FILTER_MASK_FILENAME (1 << ALLOCINFO_FILTER_FILENAME)
> #define ALLOCINFO_FILTER_MASK_LINENO (1 << ALLOCINFO_FILTER_LINENO)
> +#define ALLOCINFO_FILTER_MASK_MIN_SIZE (1 << ALLOCINFO_FILTER_MIN_SIZE)
> +#define ALLOCINFO_FILTER_MASK_MAX_SIZE (1 << ALLOCINFO_FILTER_MAX_SIZE)
>
> #define ALLOCINFO_FILTER_MASKS \
> ((1 << (__ALLOCINFO_FILTER_LAST + 1)) - 1)
> @@ -53,6 +57,8 @@ enum {
> struct allocinfo_filter {
> __u64 mask; /* bitmask of the filter fields used */
> struct allocinfo_tag fields;
> + __u64 min_size;
> + __u64 max_size;
> };
>
> struct allocinfo_get_at {
> diff --git a/lib/alloc_tag.c b/lib/alloc_tag.c
> index 56c394ef721f..6c8743eead2d 100644
> --- a/lib/alloc_tag.c
> +++ b/lib/alloc_tag.c
> @@ -173,11 +173,21 @@ static int allocinfo_cmp_str(const char *str, const char *template)
> return strncmp(allocinfo_str(str), template, ALLOCINFO_STR_SIZE);
> }
>
> +static inline struct alloc_tag_counters allocinfo_prefetch_counters(struct codetag *ct)
> +{
> + return alloc_tag_read(ct_to_alloc_tag(ct));
> +}
> +
> static void allocinfo_to_params(struct codetag *ct,
> - struct allocinfo_tag_data *data)
> + struct allocinfo_tag_data *data,
> + struct alloc_tag_counters *counters)
> {
> - struct alloc_tag *tag = ct_to_alloc_tag(ct);
> - struct alloc_tag_counters counter = alloc_tag_read(tag);
> + struct alloc_tag_counters local_counters;
> +
> + if (!counters) {
> + local_counters = allocinfo_prefetch_counters(ct);
> + counters = &local_counters;
> + }
>
> if (ct->modname)
> allocinfo_copy_str(data->tag.modname, ct->modname);
> @@ -186,9 +196,9 @@ static void allocinfo_to_params(struct codetag *ct,
> allocinfo_copy_str(data->tag.function, ct->function);
> allocinfo_copy_str(data->tag.filename, ct->filename);
> data->tag.lineno = ct->lineno;
> - data->counter.bytes = counter.bytes;
> - data->counter.calls = counter.calls;
> - data->counter.accurate = !alloc_tag_is_inaccurate(tag);
> + data->counter.bytes = counters->bytes;
> + data->counter.calls = counters->calls;
> + data->counter.accurate = !alloc_tag_is_inaccurate(ct_to_alloc_tag(ct));
> }
>
> static int allocinfo_ioctl_get_content_id(struct seq_file *m, void __user *arg)
> @@ -204,7 +214,8 @@ static int allocinfo_ioctl_get_content_id(struct seq_file *m, void __user *arg)
> return 0;
> }
>
> -static bool matches_filter(struct codetag *ct, struct allocinfo_filter *filter)
> +static bool matches_filter(struct codetag *ct, struct allocinfo_filter *filter,
> + struct alloc_tag_counters *counters)
> {
> if (!filter || !filter->mask)
> return true;
> @@ -228,6 +239,17 @@ static bool matches_filter(struct codetag *ct, struct allocinfo_filter *filter)
> ct->lineno != filter->fields.lineno)
> return false;
>
> + if ((filter->mask & ALLOCINFO_FILTER_MASK_MIN_SIZE) ||
> + (filter->mask & ALLOCINFO_FILTER_MASK_MAX_SIZE)) {
> + /* We assume counters is not NULL here as per caller logic */
> + if ((filter->mask & ALLOCINFO_FILTER_MASK_MIN_SIZE) &&
> + counters->bytes < filter->min_size)
> + return false;
> + if ((filter->mask & ALLOCINFO_FILTER_MASK_MAX_SIZE) &&
> + counters->bytes > filter->max_size)
> + return false;
> + }
> +
> return true;
> }
>
> @@ -237,6 +259,9 @@ static int allocinfo_ioctl_get_at(struct seq_file *m, void __user *arg)
> struct codetag *ct;
> struct allocinfo_get_at params = {0};
> __u64 skip_count;
> + bool sizes_set;
> + struct alloc_tag_counters counters;
> + struct alloc_tag_counters *counters_ptr = NULL;
>
> if (copy_from_user(¶ms, arg, sizeof(params)))
> return -EFAULT;
> @@ -244,9 +269,16 @@ static int allocinfo_ioctl_get_at(struct seq_file *m, void __user *arg)
> if (params.filter.mask & ~ALLOCINFO_FILTER_MASKS)
> return -EINVAL;
>
> + if ((params.filter.mask & ALLOCINFO_FILTER_MASK_MIN_SIZE) &&
> + (params.filter.mask & ALLOCINFO_FILTER_MASK_MAX_SIZE) &&
> + params.filter.min_size > params.filter.max_size)
> + return -EINVAL;
> +
> priv = (struct allocinfo_private *)m->private;
>
> skip_count = params.pos;
> + sizes_set = (params.filter.mask &
> + (ALLOCINFO_FILTER_MASK_MIN_SIZE | ALLOCINFO_FILTER_MASK_MAX_SIZE));
>
> mutex_lock(&priv->ioctl_lock);
> codetag_lock_module_list(alloc_tag_cttype, true);
> @@ -261,7 +293,11 @@ static int allocinfo_ioctl_get_at(struct seq_file *m, void __user *arg)
> ct = codetag_next_ct(&priv->ioctl_iter);
>
> while (ct) {
> - if (matches_filter(ct, &priv->filter)) {
> + if (sizes_set) {
> + counters = allocinfo_prefetch_counters(ct);
> + counters_ptr = &counters;
> + }
> + if (matches_filter(ct, &priv->filter, counters_ptr)) {
alloc_tag_read() walks all per-CPU counters which is not cheap, but here
it's called for every codetag unconditionally when sizes_set is true,
even when the tag would be rejected by modname/function/filename checks
that are plain string comparisons.
For example, say the user filters with MODNAME | MIN_SIZE on a system
with 10000 tags, 100 of which belong to the target module. Today the
code would call alloc_tag_read() 10000 times (once per tag), but only
100 of those tags pass the modname check — the other 9900 per-CPU walks
are wasted.
Would it make sense to split the filter check so that per-CPU counter reads
only happen after tag-based checks pass? Something like:
static bool allocinfo_match_tag(struct codetag *ct,
struct allocinfo_filter *filter) { ... }
static bool allocinfo_match_size(struct alloc_tag_counters *counters,
struct allocinfo_filter *filter) { ... }
And in the caller:
bool match = allocinfo_match_tag(ct, &priv->filter);
/* Add comments to help subsequent developers understand the purpose of
this modification. */
if (match && sizes_set) {
counters = allocinfo_prefetch_counters(ct);
counters_ptr = &counters;
match = allocinfo_match_size(counters_ptr, &priv->filter);
}
You may find a more elegant approach to resolve this issue.
Thanks
Best Regards
Hao
> if (skip_count == 0)
> break;
> skip_count--;
> @@ -270,7 +306,7 @@ static int allocinfo_ioctl_get_at(struct seq_file *m, void __user *arg)
> }
>
> if (ct) {
> - allocinfo_to_params(ct, ¶ms.data);
> + allocinfo_to_params(ct, ¶ms.data, counters_ptr);
> priv->positioned = true;
> }
>
> @@ -292,9 +328,15 @@ static int allocinfo_ioctl_get_next(struct seq_file *m, void __user *arg)
> struct codetag *ct;
> struct allocinfo_tag_data params = {0};
> int ret = 0;
> + bool sizes_set;
> + struct alloc_tag_counters counters;
> + struct alloc_tag_counters *counters_ptr = NULL;
>
> priv = (struct allocinfo_private *)m->private;
>
> + sizes_set = (priv->filter.mask &
> + (ALLOCINFO_FILTER_MASK_MIN_SIZE | ALLOCINFO_FILTER_MASK_MAX_SIZE));
> +
> mutex_lock(&priv->ioctl_lock);
> codetag_lock_module_list(alloc_tag_cttype, true);
>
> @@ -304,10 +346,18 @@ static int allocinfo_ioctl_get_next(struct seq_file *m, void __user *arg)
> }
>
> ct = codetag_next_ct(&priv->ioctl_iter);
> - while (ct && !matches_filter(ct, &priv->filter))
> + while (ct) {
> + if (sizes_set) {
> + counters = allocinfo_prefetch_counters(ct);
> + counters_ptr = &counters;
> + }
> + if (matches_filter(ct, &priv->filter, counters_ptr))
> + break;
> ct = codetag_next_ct(&priv->ioctl_iter);
> + }
> +
> if (ct)
> - allocinfo_to_params(ct, ¶ms);
> + allocinfo_to_params(ct, ¶ms, counters_ptr);
>
> if (!ct) {
> priv->positioned = false;
next prev parent reply other threads:[~2026-05-26 3:12 UTC|newest]
Thread overview: 13+ messages / expand[flat|nested] mbox.gz Atom feed top
2026-05-22 17:45 [PATCH v2 0/6] alloc_tag: introduce IOCTL-based filtering for MAP Abhishek Bapat
2026-05-22 17:45 ` [PATCH v2 1/6] alloc_tag: add ioctl to /proc/allocinfo Abhishek Bapat
2026-05-22 20:11 ` Andrew Morton
2026-05-25 2:20 ` Hao Ge
2026-05-22 17:45 ` [PATCH v2 2/6] alloc_tag: add ioctl filters " Abhishek Bapat
2026-05-25 2:59 ` Hao Ge
2026-05-22 17:45 ` [PATCH v2 3/6] alloc_tag: add size-based filtering to ioctl Abhishek Bapat
2026-05-26 3:11 ` Hao Ge [this message]
2026-05-22 17:45 ` [PATCH v2 4/6] alloc_tag: add accuracy based " Abhishek Bapat
2026-05-22 17:45 ` [PATCH v2 5/6] kselftest: alloc_tag: add kselftest for ioctl interface Abhishek Bapat
2026-05-22 17:45 ` [PATCH v2 6/6] kselftest: alloc_tag: extend the allocinfo ioctl kselftest Abhishek Bapat
2026-05-22 20:11 ` [PATCH v2 0/6] alloc_tag: introduce IOCTL-based filtering for MAP Andrew Morton
2026-05-25 7:32 ` Hao Ge
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=de0f2984-44ea-4098-9d19-c63ee035cdaf@linux.dev \
--to=hao.ge@linux.dev \
--cc=abhishekbapat@google.com \
--cc=akpm@linux-foundation.org \
--cc=corbet@lwn.net \
--cc=kent.overstreet@linux.dev \
--cc=linux-doc@vger.kernel.org \
--cc=linux-kernel@vger.kernel.org \
--cc=linux-mm@kvack.org \
--cc=skhan@linuxfoundation.org \
--cc=souravpanda@google.com \
--cc=surenb@google.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