[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Message-ID: <aTDjmdfUk94R3Dyk@lizhi-Precision-Tower-5810>
Date: Wed, 3 Dec 2025 20:27:53 -0500
From: Frank Li <Frank.li@....com>
To: Dave Jiang <dave.jiang@...el.com>
Cc: Vinod Koul <vkoul@...nel.org>, Dong Aisheng <aisheng.dong@....com>,
Andi Shyti <andi.shyti@...nel.org>, Shawn Guo <shawnguo@...nel.org>,
Sascha Hauer <s.hauer@...gutronix.de>,
Pengutronix Kernel Team <kernel@...gutronix.de>,
Fabio Estevam <festevam@...il.com>, dmaengine@...r.kernel.org,
linux-kernel@...r.kernel.org, linux-i2c@...r.kernel.org,
imx@...ts.linux.dev, linux-arm-kernel@...ts.infradead.org,
carlos.song@....com
Subject: Re: [PATCH v2 2/2] i2c: lpi2c: Use cleanup helper for
dma_async_tx_descriptor error handling
On Wed, Dec 03, 2025 at 03:52:09PM -0700, Dave Jiang wrote:
>
>
> On 10/3/25 9:26 AM, Frank Li wrote:
> > Use the cleanup helper to simplify dma_async_tx_descriptor error handling.
> >
> > No functional change.
> >
> > Signed-off-by: Frank Li <Frank.Li@....com>
> > ---
> > change v2
> > need goto unmap buffer.
> >
> > desc_prepare_err_exit:
> > dma_unmap_single(txchan->device->dev, dma->dma_tx_addr,
> > dma->rx_cmd_buf_len, DMA_TO_DEVICE);
> > return -EINVAL;
> >
> > submit_err_exit:
> > dma_unmap_single(txchan->device->dev, dma->dma_tx_addr,
> > dma->rx_cmd_buf_len, DMA_TO_DEVICE);
> > dmaengine_desc_free(rx_cmd_desc)
> >
> > auto cleanup by __free, so whole error path submit_err_exit can be removed
> > and use desc_prepare_err_exit label.
> > ---
> > drivers/i2c/busses/i2c-imx-lpi2c.c | 11 +++--------
> > 1 file changed, 3 insertions(+), 8 deletions(-)
> >
> > diff --git a/drivers/i2c/busses/i2c-imx-lpi2c.c b/drivers/i2c/busses/i2c-imx-lpi2c.c
> > index 03b5a7e8c361abe1d75fb4d31f9614bbc6387d93..e31c2d81f59f18a09d34f79896a7fa4bb83c0a12 100644
> > --- a/drivers/i2c/busses/i2c-imx-lpi2c.c
> > +++ b/drivers/i2c/busses/i2c-imx-lpi2c.c
> > @@ -722,7 +722,7 @@ static void lpi2c_dma_callback(void *data)
> >
> > static int lpi2c_dma_rx_cmd_submit(struct lpi2c_imx_struct *lpi2c_imx)
> > {
> > - struct dma_async_tx_descriptor *rx_cmd_desc;
> > + struct dma_async_tx_descriptor *rx_cmd_desc __free(dma_async_tx_descriptor) = NULL;
> > struct lpi2c_imx_dma *dma = lpi2c_imx->dma;
> > struct dma_chan *txchan = dma->chan_tx;
> > dma_cookie_t cookie;
> > @@ -746,10 +746,11 @@ static int lpi2c_dma_rx_cmd_submit(struct lpi2c_imx_struct *lpi2c_imx)
> > cookie = dmaengine_submit(rx_cmd_desc);
> > if (dma_submit_error(cookie)) {
> > dev_err(&lpi2c_imx->adapter.dev, "submitting DMA failed, use pio\n");
> > - goto submit_err_exit;
> > + goto desc_prepare_err_exit;
>
> Typically if you are introducing cleanup usage, you will want to remove all the goto usages. Otherwise it gets kinda messy having both styles. Although not sure how to do a cleanup macro with dma_unmap_*. Unless a local one where lpi2c_imx_dma also contains a ptr to the device. Then that may be doable.
dma_unmap() patch at
https://lore.kernel.org/imx/aPZbuY5bLpkoO2J%2F@lizhi-Precision-Tower-5810/
The first patch need be pickup and dma umap firstly. Then I can remove all
goto.
Frank
>
> DJ> }
> >
> > dma_async_issue_pending(txchan);
> > + retain_and_null_ptr(rx_cmd_desc);
> >
> > return 0;
> >
> > @@ -757,12 +758,6 @@ static int lpi2c_dma_rx_cmd_submit(struct lpi2c_imx_struct *lpi2c_imx)
> > dma_unmap_single(txchan->device->dev, dma->dma_tx_addr,
> > dma->rx_cmd_buf_len, DMA_TO_DEVICE);
> > return -EINVAL;
> > -
> > -submit_err_exit:
> > - dma_unmap_single(txchan->device->dev, dma->dma_tx_addr,
> > - dma->rx_cmd_buf_len, DMA_TO_DEVICE);
> > - dmaengine_desc_free(rx_cmd_desc);
> > - return -EINVAL;
> > }
> >
> > static int lpi2c_dma_submit(struct lpi2c_imx_struct *lpi2c_imx)
> >
>
Powered by blists - more mailing lists