On Thu, 2015-09-17 at 16:47 -0500, Pierre-Louis Bossart wrote:
> +#if LINUX_VERSION_CODE < KERNEL_VERSION(4,2,0)
> +
> +#define hrtimer_resolution (unsigned int)LOW_RES_NSEC
> +
> +#endif /* < 4.2 */
Do we need parentheses around that all?
> --- a/backport/backport-include/linux/skbuff.h
> +++ b/backport/backport-include/linux/skbuff.h
> @@ -258,6 +258,8 @@ static inline struct page *dev_alloc_page(void)
> #endif /* LINUX_VERSION_CODE < KERNEL_VERSION(3,19,0) */
>
> #if LINUX_VERSION_CODE < KERNEL_VERSION(3,19,0)
> +
> +#include <linux/uio.h>
That's a bit odd? But perhaps necessary - just better to have the blank
line after the include :)
> --- /dev/null
> +++ b/backport/backport-include/linux/uio.h
> @@ -0,0 +1,246 @@
> +#if (LINUX_VERSION_CODE >= KERNEL_VERSION(3,5,0))
> +#include_next <linux/uio.h>
> +
> +#ifndef __iter_is_iovec_h_
> +#define __iter_is_iovec_h_
Perhaps better to simply protect the whole file?
> +#if (LINUX_VERSION_CODE < KERNEL_VERSION(4,2,0))
> +
> +#if (LINUX_VERSION_CODE < KERNEL_VERSION(3,15,0))
> +
> +/* iov_iter is defined elsewhere */
> +#include_next <linux/fs.h>
> +static inline bool iter_is_iovec(struct iov_iter *i)
> +{
> + return true;
> +}
> +
> +#else
> +static inline bool iter_is_iovec(struct iov_iter *i)
> +{
> + return !(i->type & (ITER_BVEC | ITER_KVEC));
> +}
> +#endif
> +#endif
That could be written a bit clearer with #elif, I think.
> +#endif /* __iter_is_iovec_h_ */
> +
> +#else
> +
> +#ifndef _LINUX_UIO_H
> +#define _LINUX_UIO_H
>
This part looks like a straight-up copy - perhaps better to add it to
copy-list and then provide a patch to patch around it the necessary
modifications - see for example the devcoredump patch in
patches/backport-adjustments/devcoredump.patch
> --- a/dependencies
> +++ b/dependencies
> @@ -178,3 +178,5 @@ IGB 3.5
>
> # This driver needs mmc_hw_reset() which was added in kernel version
> 3.2
> MWIFIEX_SDIO 3.2
> +
> +SOUND 3.13
> \ No newline at end of file
Please keep a newline at EOF.
> ++static ssize_t snd_pcm_aio_read(struct kiocb *iocb, const struct
> iovec *iov,
> ++ unsigned long nr_segs, loff_t pos)
> ++
Is it possible to reduce this function to calling the existing
snd_pcm_readv (and aio_write to writev) by suitable data structure
manipulation?
If so, it might be possible to create an spatch to make these changes,
which would make them apply more generally and be less prone to
breaking.
johannes
--
To unsubscribe from this list: send the line "unsubscribe backports" in
