Re: [RFC PATCH] block/export/fuse: Fix build failure on FreeBSD

[Kevin Wolf]

Am 22.01.2022 um 14:49 hat Philippe Mathieu-Daudé geschrieben:
> When building on FreeBSD we get:
> 
>   [816/6851] Compiling C object libblockdev.fa.p/block_export_fuse.c.o
>   ../block/export/fuse.c:628:16: error: use of undeclared identifier 
> 'FALLOC_FL_KEEP_SIZE'
>       if (mode & FALLOC_FL_KEEP_SIZE) {
>                  ^
>   ../block/export/fuse.c:632:16: error: use of undeclared identifier 
> 'FALLOC_FL_PUNCH_HOLE'
>       if (mode & FALLOC_FL_PUNCH_HOLE) {
>                  ^
>   ../block/export/fuse.c:633:22: error: use of undeclared identifier 
> 'FALLOC_FL_KEEP_SIZE'
>           if (!(mode & FALLOC_FL_KEEP_SIZE)) {
>                        ^
>   3 errors generated.
>   FAILED: libblockdev.fa.p/block_export_fuse.c.o
> 
> Meson indeed reported FALLOC_FL_PUNCH_HOLE is not available:
> 
>   C compiler for the host machine: cc (clang 10.0.1 "FreeBSD clang version 
> 10.0.1")
>   Checking for function "fallocate" : NO
>   Checking for function "posix_fallocate" : YES
>   Header <linux/falloc.h> has symbol "FALLOC_FL_PUNCH_HOLE" : NO
>   Header <linux/falloc.h> has symbol "FALLOC_FL_ZERO_RANGE" : NO
>   ...
> 
> Similarly to commit 304332039 ("block/export/fuse.c: fix musl build"),
> guard the code requiring FALLOC_FL_KEEP_SIZE / FALLOC_FL_PUNCH_HOLE
> definitions under CONFIG_FALLOCATE_PUNCH_HOLE #ifdef'ry.
> 
> Signed-off-by: Philippe Mathieu-Daudé <f4bug@amsat.org>
> ---
> Fragile #ifdef'ry... Any clever idea?

I guess we could reorder things. The last case (!mode) is mutually
exclusive with the other conditions, so checking it first doesn't make a
difference, and then you can #ifdef things out more cleanly.

>     ERROR: else should follow close brace '}'
>     #17: FILE: block/export/fuse.c:647:
>          }
>     +    else
> 
>     ERROR: else should follow close brace '}'
>     #29: FILE: block/export/fuse.c:670:
>          }
>     +    else
> 
>     total: 2 errors, 0 warnings, 28 lines checked

On the other hand, I'm not really worried about these either. The only
way this could fail in the future is build errors, which are obvious and
quick to fix.

If you hadn't marked this an RFC, I think I would just have applied it.

Kevin

> ---
>  block/export/fuse.c | 8 ++++++--
>  1 file changed, 6 insertions(+), 2 deletions(-)
> 
> diff --git a/block/export/fuse.c b/block/export/fuse.c
> index 6710d8aed86..d8bad0e53df 100644
> --- a/block/export/fuse.c
> +++ b/block/export/fuse.c
> @@ -625,6 +625,7 @@ static void fuse_fallocate(fuse_req_t req, fuse_ino_t 
> inode, int mode,
>          return;
>      }
>  
> +#ifdef CONFIG_FALLOCATE_PUNCH_HOLE
>      if (mode & FALLOC_FL_KEEP_SIZE) {
>          length = MIN(length, blk_len - offset);
>      }
> @@ -643,8 +644,10 @@ static void fuse_fallocate(fuse_req_t req, fuse_ino_t 
> inode, int mode,
>              length -= size;
>          } while (ret == 0 && length > 0);
>      }
> +    else
> +#endif /* CONFIG_FALLOCATE_PUNCH_HOLE */
>  #ifdef CONFIG_FALLOCATE_ZERO_RANGE
> -    else if (mode & FALLOC_FL_ZERO_RANGE) {
> +    if (mode & FALLOC_FL_ZERO_RANGE) {
>          if (!(mode & FALLOC_FL_KEEP_SIZE) && offset + length > blk_len) {
>              /* No need for zeroes, we are going to write them ourselves */
>              ret = fuse_do_truncate(exp, offset + length, false,
> @@ -664,8 +667,9 @@ static void fuse_fallocate(fuse_req_t req, fuse_ino_t 
> inode, int mode,
>              length -= size;
>          } while (ret == 0 && length > 0);
>      }
> +    else
>  #endif /* CONFIG_FALLOCATE_ZERO_RANGE */
> -    else if (!mode) {
> +    if (!mode) {
>          /* We can only fallocate at the EOF with a truncate */
>          if (offset < blk_len) {
>              fuse_reply_err(req, EOPNOTSUPP);
> -- 
> 2.34.1
>