Re: [PATCH 9/9 v3] spi: add Broadcom BCM63xx SPI controller driver
From: Grant Likely <hidden>
Date: 2012-01-31 21:59:14
Also in:
linux-spi
Possibly related (same subject, not in this thread)
- 2012-01-31 · Re: [PATCH 9/9 v3] spi: add Broadcom BCM63xx SPI controller driver · Florian Fainelli <hidden>
- 2012-01-31 · Re: [PATCH 9/9 v3] spi: add Broadcom BCM63xx SPI controller driver · Grant Likely <hidden>
- 2012-01-31 · Re: [PATCH 9/9 v3] spi: add Broadcom BCM63xx SPI controller driver · Shubhrajyoti Datta <hidden>
- 2012-01-31 · [PATCH 9/9 v3] spi: add Broadcom BCM63xx SPI controller driver · Florian Fainelli <florian@openwrt.org>
On Tue, Jan 31, 2012 at 2:20 PM, Florian Fainelli [off-list ref] wrote:
On Tuesday 31 January 2012 21:19:22 Grant Likely wrote:quoted
quoted
+static const struct dev_pm_ops bcm63xx_spi_pm_ops = { + .suspend = bcm63xx_spi_suspend, + .resume = bcm63xx_spi_resume, +}; + +#define BCM63XX_SPI_PM_OPS (&bcm63xx_spi_pm_ops) +#else +#define BCM63XX_SPI_PM_OPS NULLA bit ugly. Do this instead in the else clause and drop the BCM63XX_SPI_PM_OPS: #define bcm63xx_spi_pm_ops NULLThis won't work, because driver.pm must be set to a pointer to a struct dev_pm_ops, that's why I used this trick to make it build fine in both cases. If I follow your advice, with driver.pm = &bcm63xx_spi_pm_ops, it won't build for CONFIG_PM=n.
Okay, fair enough. g.