DPDK patches and discussions
 help / color / mirror / Atom feed
From: bugzilla@dpdk.org
To: dev@dpdk.org
Subject: [DPDK/core Bug 1525] dpdk mempool creation is resulting into lot of memory getting wasted
Date: Wed, 21 Aug 2024 10:57:07 +0000	[thread overview]
Message-ID: <bug-1525-3@http.bugs.dpdk.org/> (raw)

[-- Attachment #1: Type: text/plain, Size: 1636 bytes --]

https://bugs.dpdk.org/show_bug.cgi?id=1525

            Bug ID: 1525
           Summary: dpdk mempool creation is resulting into lot of memory
                    getting wasted
           Product: DPDK
           Version: 20.11
          Hardware: x86
                OS: Linux
            Status: UNCONFIRMED
          Severity: major
          Priority: Normal
         Component: core
          Assignee: dev@dpdk.org
          Reporter: sandipchhikara@gmail.com
  Target Milestone: ---

We are using rte_mempool_op_populate_helper for creating dpdk mempool.  Memory
is allocated from 1GB huge page and passed to this function. It is observed
that rte_mempool_op_populate_helper  function is calculating page size as 4K
and as MEMPOOL_F_NO_IOVA_CONTIG is not specified, so it don't place objects
across pages resulting into lot of memory wastage and creation of pool which is
lesser in size then required.
rte_mempool_op_calc_mem_size_default is used to calculate how much memory is
required for the pool. 

If it is user application, which is working on virtual memory (No driver
accessing physical memory directly), whether it would be fine to enable flag
MEMPOOL_F_NO_IOVA_CONTIG. 

We have object of 712 bytes, (final memory for each object is becoming 812
bytes after accounting header and trailer etc).
We are creating 2500000 objects in pool, allocating 2 GB (1 GB huge pages)
memory for the same, but after placing 4 objects in one page, it starts using
next page as it is wrongly finding that page size is 4K.

-- 
You are receiving this mail because:
You are the assignee for the bug.

[-- Attachment #2: Type: text/html, Size: 3511 bytes --]

                 reply	other threads:[~2024-08-21 10:57 UTC|newest]

Thread overview: [no followups] expand[flat|nested]  mbox.gz  Atom feed

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=bug-1525-3@http.bugs.dpdk.org/ \
    --to=bugzilla@dpdk.org \
    --cc=dev@dpdk.org \
    /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;
as well as URLs for NNTP newsgroup(s).