From: Christophe Leroy <christophe.leroy@csgroup.eu>
To: Rasmus Villemoes <rasmus.villemoes@prevas.dk>,
Li Yang <leoyang.li@nxp.com>,
"David S. Miller" <davem@davemloft.net>,
Jakub Kicinski <kuba@kernel.org>
Cc: Vladimir Oltean <vladimir.oltean@nxp.com>,
linux-kernel@vger.kernel.org, netdev@vger.kernel.org,
linuxppc-dev@lists.ozlabs.org, Zhao Qiang <qiang.zhao@nxp.com>
Subject: Re: [PATCH 14/20] ethernet: ucc_geth: don't statically allocate eight ucc_geth_info
Date: Tue, 8 Dec 2020 16:13:15 +0100 [thread overview]
Message-ID: <8259bec3-9343-82e3-a420-a8170cf922a4@csgroup.eu> (raw)
In-Reply-To: <20201205191744.7847-15-rasmus.villemoes@prevas.dk>
Le 05/12/2020 à 20:17, Rasmus Villemoes a écrit :
> struct ucc_geth_info is somewhat large, and on systems with only one
> or two UCC instances, that just wastes a few KB of memory. So
> allocate and populate a chunk of memory at probe time instead of
> initializing them all during driver init.
>
> Note that the existing "ug_info == NULL" check was dead code, as the
> address of some static array element can obviously never be NULL.
>
> Signed-off-by: Rasmus Villemoes <rasmus.villemoes@prevas.dk>
> ---
> drivers/net/ethernet/freescale/ucc_geth.c | 32 +++++++++--------------
> 1 file changed, 12 insertions(+), 20 deletions(-)
>
> diff --git a/drivers/net/ethernet/freescale/ucc_geth.c b/drivers/net/ethernet/freescale/ucc_geth.c
> index a06744d8b4af..273342233bba 100644
> --- a/drivers/net/ethernet/freescale/ucc_geth.c
> +++ b/drivers/net/ethernet/freescale/ucc_geth.c
> @@ -157,8 +157,6 @@ static const struct ucc_geth_info ugeth_primary_info = {
> .riscRx = QE_RISC_ALLOCATION_RISC1_AND_RISC2,
> };
>
> -static struct ucc_geth_info ugeth_info[8];
> -
> #ifdef DEBUG
> static void mem_disp(u8 *addr, int size)
> {
> @@ -3714,25 +3712,23 @@ static int ucc_geth_probe(struct platform_device* ofdev)
> if ((ucc_num < 0) || (ucc_num > 7))
> return -ENODEV;
>
> - ug_info = &ugeth_info[ucc_num];
> - if (ug_info == NULL) {
> - if (netif_msg_probe(&debug))
> - pr_err("[%d] Missing additional data!\n", ucc_num);
> - return -ENODEV;
> - }
> + ug_info = kmalloc(sizeof(*ug_info), GFP_KERNEL);
Could we use dev_kmalloc() instead, to avoid the freeing on the wait out and the err_free_info: path ?
> + if (ug_info == NULL)
> + return -ENOMEM;
> + memcpy(ug_info, &ugeth_primary_info, sizeof(*ug_info));
>
> ug_info->uf_info.ucc_num = ucc_num;
>
> err = ucc_geth_parse_clock(np, "rx", &ug_info->uf_info.rx_clock);
> if (err)
> - return err;
> + goto err_free_info;
> err = ucc_geth_parse_clock(np, "tx", &ug_info->uf_info.tx_clock);
> if (err)
> - return err;
> + goto err_free_info;
>
> err = of_address_to_resource(np, 0, &res);
> if (err)
> - return -EINVAL;
> + goto err_free_info;
>
> ug_info->uf_info.regs = res.start;
> ug_info->uf_info.irq = irq_of_parse_and_map(np, 0);
> @@ -3745,7 +3741,7 @@ static int ucc_geth_probe(struct platform_device* ofdev)
> */
> err = of_phy_register_fixed_link(np);
> if (err)
> - return err;
> + goto err_free_info;
> ug_info->phy_node = of_node_get(np);
> }
>
> @@ -3876,6 +3872,8 @@ static int ucc_geth_probe(struct platform_device* ofdev)
> of_phy_deregister_fixed_link(np);
> of_node_put(ug_info->tbi_node);
> of_node_put(ug_info->phy_node);
> +err_free_info:
> + kfree(ug_info);
>
> return err;
> }
> @@ -3886,6 +3884,7 @@ static int ucc_geth_remove(struct platform_device* ofdev)
> struct ucc_geth_private *ugeth = netdev_priv(dev);
> struct device_node *np = ofdev->dev.of_node;
>
> + kfree(ugeth->ug_info);
> ucc_geth_memclean(ugeth);
> if (of_phy_is_fixed_link(np))
> of_phy_deregister_fixed_link(np);
> @@ -3920,17 +3919,10 @@ static struct platform_driver ucc_geth_driver = {
>
> static int __init ucc_geth_init(void)
> {
> - int i, ret;
> -
> if (netif_msg_drv(&debug))
> pr_info(DRV_DESC "\n");
> - for (i = 0; i < 8; i++)
> - memcpy(&(ugeth_info[i]), &ugeth_primary_info,
> - sizeof(ugeth_primary_info));
> -
> - ret = platform_driver_register(&ucc_geth_driver);
>
> - return ret;
> + return platform_driver_register(&ucc_geth_driver);
> }
>
> static void __exit ucc_geth_exit(void)
>
next prev parent reply other threads:[~2020-12-08 15:14 UTC|newest]
Thread overview: 46+ messages / expand[flat|nested] mbox.gz Atom feed top
2020-12-05 19:17 [PATCH 00/20] ethernet: ucc_geth: assorted fixes and simplifications Rasmus Villemoes
2020-12-05 19:17 ` [PATCH 01/20] ethernet: ucc_geth: set dev->max_mtu to 1518 Rasmus Villemoes
2020-12-10 1:25 ` Andrew Lunn
2021-01-05 14:17 ` Joakim Tjernlund
2021-01-05 14:33 ` Andrew Lunn
2021-01-05 14:44 ` Joakim Tjernlund
2021-01-05 14:54 ` Andrew Lunn
2020-12-05 19:17 ` [PATCH 02/20] ethernet: ucc_geth: fix definition and size of ucc_geth_tx_global_pram Rasmus Villemoes
2020-12-08 19:14 ` Li Yang
2020-12-08 20:12 ` Rasmus Villemoes
2020-12-05 19:17 ` [PATCH 03/20] ethernet: ucc_geth: remove unused read of temoder field Rasmus Villemoes
2020-12-05 19:17 ` [PATCH 04/20] soc: fsl: qe: make cpm_muram_offset take a const void* argument Rasmus Villemoes
2020-12-05 19:17 ` [PATCH 05/20] soc: fsl: qe: store muram_vbase as a void pointer instead of u8 Rasmus Villemoes
2020-12-05 19:17 ` [PATCH 06/20] soc: fsl: qe: add cpm_muram_free_addr() helper Rasmus Villemoes
2020-12-05 19:17 ` [PATCH 07/20] ethernet: ucc_geth: use qe_muram_free_addr() Rasmus Villemoes
2020-12-10 1:36 ` Andrew Lunn
2020-12-05 19:17 ` [PATCH 08/20] ethernet: ucc_geth: remove unnecessary memset_io() calls Rasmus Villemoes
2020-12-05 19:17 ` [PATCH 09/20] ethernet: ucc_geth: replace kmalloc+memset by kzalloc Rasmus Villemoes
2020-12-05 19:17 ` [PATCH 10/20] ethernet: ucc_geth: remove {rx,tx}_glbl_pram_offset from struct ucc_geth_private Rasmus Villemoes
2020-12-05 19:17 ` [PATCH 11/20] ethernet: ucc_geth: fix use-after-free in ucc_geth_remove() Rasmus Villemoes
2020-12-05 20:48 ` Jakub Kicinski
2020-12-05 21:04 ` Rasmus Villemoes
2020-12-05 21:19 ` Jakub Kicinski
2020-12-05 21:35 ` Rasmus Villemoes
2020-12-05 21:50 ` Rasmus Villemoes
2020-12-05 19:17 ` [PATCH 12/20] ethernet: ucc_geth: factor out parsing of {rx,tx}-clock{,-name} properties Rasmus Villemoes
2020-12-05 19:17 ` [PATCH 13/20] ethernet: ucc_geth: constify ugeth_primary_info Rasmus Villemoes
2020-12-05 19:17 ` [PATCH 14/20] ethernet: ucc_geth: don't statically allocate eight ucc_geth_info Rasmus Villemoes
2020-12-08 15:13 ` Christophe Leroy [this message]
2020-12-08 21:17 ` Rasmus Villemoes
2020-12-05 19:17 ` [PATCH 15/20] ethernet: ucc_geth: use UCC_GETH_{RX,TX}_BD_RING_ALIGNMENT macros directly Rasmus Villemoes
2020-12-08 15:14 ` [PATCH 15/20] ethernet: ucc_geth: use UCC_GETH_{RX, TX}_BD_RING_ALIGNMENT " Christophe Leroy
2020-12-05 19:17 ` [PATCH 16/20] ethernet: ucc_geth: remove bd_mem_part and all associated code Rasmus Villemoes
2020-12-05 19:17 ` [PATCH 17/20] ethernet: ucc_geth: replace kmalloc_array()+for loop by kcalloc() Rasmus Villemoes
2020-12-05 19:17 ` [PATCH 18/20] ethernet: ucc_geth: add helper to replace repeated switch statements Rasmus Villemoes
2020-12-08 15:21 ` Christophe Leroy
2020-12-08 20:55 ` Rasmus Villemoes
2020-12-05 19:17 ` [PATCH 19/20] ethernet: ucc_geth: inform the compiler that numQueues is always 1 Rasmus Villemoes
2020-12-05 19:17 ` [PATCH 20/20] ethernet: ucc_geth: simplify rx/tx allocations Rasmus Villemoes
2020-12-05 20:53 ` [PATCH 00/20] ethernet: ucc_geth: assorted fixes and simplifications Jakub Kicinski
2020-12-05 21:11 ` Rasmus Villemoes
2020-12-05 21:27 ` Jakub Kicinski
2020-12-05 21:36 ` Rasmus Villemoes
2020-12-10 7:52 ` Rasmus Villemoes
2020-12-08 3:07 ` Qiang Zhao
2020-12-08 8:13 ` Rasmus Villemoes
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=8259bec3-9343-82e3-a420-a8170cf922a4@csgroup.eu \
--to=christophe.leroy@csgroup.eu \
--cc=davem@davemloft.net \
--cc=kuba@kernel.org \
--cc=leoyang.li@nxp.com \
--cc=linux-kernel@vger.kernel.org \
--cc=linuxppc-dev@lists.ozlabs.org \
--cc=netdev@vger.kernel.org \
--cc=qiang.zhao@nxp.com \
--cc=rasmus.villemoes@prevas.dk \
--cc=vladimir.oltean@nxp.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;
as well as URLs for NNTP newsgroup(s).