Even when a PMD was disabled with meson's disable_drivers option its config file was still being parsed. Some of the PMD configs attempt to find a library they depend on and parse its header files with certain assumptions. If the library is found, but it's simply too old to contain the necessary header files, the meson build fails and it can only be fixed by either updating that library, or expanding the meson script for the faulty PMD. While the latter should be still done for the sake of DPDK quality, an intermediate solution would be to skip building the faulty PMD - there's a chance we don't need it. That's what this patch allows. Signed-off-by: Darek Stojaczyk <dariusz.stojaczyk@intel.com> --- drivers/meson.build | 9 ++++++--- 1 file changed, 6 insertions(+), 3 deletions(-) diff --git a/drivers/meson.build b/drivers/meson.build index 5502bf9924..a13c62a3b0 100644 --- a/drivers/meson.build +++ b/drivers/meson.build @@ -60,9 +60,6 @@ foreach class:dpdk_driver_classes ext_deps = [] pkgconfig_extra_libs = [] - # pull in driver directory which should assign to each of the above - subdir(drv_path) - # skip disabled drivers. For meson 0.49 change this to use # "in" keyword foreach disable_path: disabled_drivers @@ -71,6 +68,12 @@ foreach class:dpdk_driver_classes reason = 'Explicitly disabled via build config' endif endforeach + + if build + # pull in driver directory which should update all the local variables + subdir(drv_path) + endif + if build # get dependency objs from strings shared_deps = ext_deps -- 2.17.1
On Thu, Mar 26, 2020 at 10:22:59AM +0100, Darek Stojaczyk wrote:
> Even when a PMD was disabled with meson's disable_drivers option
> its config file was still being parsed. Some of the PMD configs
> attempt to find a library they depend on and parse its header files
> with certain assumptions. If the library is found, but it's simply
> too old to contain the necessary header files, the meson build
> fails and it can only be fixed by either updating that library, or
> expanding the meson script for the faulty PMD.
>
> While the latter should be still done for the sake of DPDK quality,
> an intermediate solution would be to skip building the faulty PMD
> - there's a chance we don't need it. That's what this patch allows.
>
> Signed-off-by: Darek Stojaczyk <dariusz.stojaczyk@intel.com>
> ---
> drivers/meson.build | 9 ++++++---
> 1 file changed, 6 insertions(+), 3 deletions(-)
>
> diff --git a/drivers/meson.build b/drivers/meson.build
> index 5502bf9924..a13c62a3b0 100644
> --- a/drivers/meson.build
> +++ b/drivers/meson.build
> @@ -60,9 +60,6 @@ foreach class:dpdk_driver_classes
> ext_deps = []
> pkgconfig_extra_libs = []
>
> - # pull in driver directory which should assign to each of the above
> - subdir(drv_path)
> -
> # skip disabled drivers. For meson 0.49 change this to use
> # "in" keyword
> foreach disable_path: disabled_drivers
> @@ -71,6 +68,12 @@ foreach class:dpdk_driver_classes
> reason = 'Explicitly disabled via build config'
> endif
> endforeach
> +
> + if build
> + # pull in driver directory which should update all the local variables
> + subdir(drv_path)
> + endif
> +
> if build
> # get dependency objs from strings
> shared_deps = ext_deps
> --
Good idea to fix this. The original idea was to parse the subdir files so
that the "reason" field could report out on any missing dependencies even
if the driver was explicitly disabled. However, that choice did cause other
problems so I think skipping the parsing as you do here is the better
choice.
Acked-by: Bruce Richardson <bruce.richardson@intel.com>
Note: if for a future release we bump our minimum meson version to 0.49 or
greater we should be able to simplify all these loops and "if build" checks
as the "break" and "continue" keywords become available, as does the "in"
keyword for checking for a value in an array.
On Thu, Mar 26, 2020 at 10:22:59AM +0100, Darek Stojaczyk wrote:
> Even when a PMD was disabled with meson's disable_drivers option
> its config file was still being parsed. Some of the PMD configs
> attempt to find a library they depend on and parse its header files
> with certain assumptions. If the library is found, but it's simply
> too old to contain the necessary header files, the meson build
> fails and it can only be fixed by either updating that library, or
> expanding the meson script for the faulty PMD.
>
> While the latter should be still done for the sake of DPDK quality,
> an intermediate solution would be to skip building the faulty PMD
> - there's a chance we don't need it. That's what this patch allows.
>
> Signed-off-by: Darek Stojaczyk <dariusz.stojaczyk@intel.com>
> ---
> drivers/meson.build | 9 ++++++---
> 1 file changed, 6 insertions(+), 3 deletions(-)
>
> diff --git a/drivers/meson.build b/drivers/meson.build
> index 5502bf9924..a13c62a3b0 100644
> --- a/drivers/meson.build
> +++ b/drivers/meson.build
> @@ -60,9 +60,6 @@ foreach class:dpdk_driver_classes
> ext_deps = []
> pkgconfig_extra_libs = []
>
> - # pull in driver directory which should assign to each of the above
> - subdir(drv_path)
> -
> # skip disabled drivers. For meson 0.49 change this to use
> # "in" keyword
> foreach disable_path: disabled_drivers
> @@ -71,6 +68,12 @@ foreach class:dpdk_driver_classes
> reason = 'Explicitly disabled via build config'
> endif
> endforeach
> +
> + if build
> + # pull in driver directory which should update all the local variables
> + subdir(drv_path)
> + endif
> +
Looking at this code and the meson docs again, I think this block and
previous can be simplified by using the array "contains()" method to avoid
the loop.
if disabled_drivers.contains(drv_path)
build = false
reson = '....'
else
# pull in driver ...
subdir(drv_path)
endif
Regards,
/Bruce
Even when a PMD was disabled with meson's disable_drivers option its config file was still being parsed. Some of the PMD configs attempt to find a library they depend on and parse its header files with certain assumptions. If the library is found, but it's simply too old to contain the necessary header files, the meson build fails and it can only be fixed by either updating that library, or expanding the meson script for the faulty PMD. While the latter should be still done for the sake of DPDK quality, an intermediate solution would be to skip building the faulty PMD - there's a chance we don't need it. That's what this patch allows. Signed-off-by: Darek Stojaczyk <dariusz.stojaczyk@intel.com> Acked-by: Bruce Richardson <bruce.richardson@intel.com> --- v2: - simplified the code with if X.contains(), as Bruce suggested - removed the related comment from the code - it should be self-documented now drivers/meson.build | 19 ++++++++----------- 1 file changed, 8 insertions(+), 11 deletions(-) diff --git a/drivers/meson.build b/drivers/meson.build index 72eec46088..483226a789 100644 --- a/drivers/meson.build +++ b/drivers/meson.build @@ -57,17 +57,14 @@ foreach class:dpdk_driver_classes ext_deps = [] pkgconfig_extra_libs = [] - # pull in driver directory which should assign to each of the above - subdir(drv_path) - - # skip disabled drivers. For meson 0.49 change this to use - # "in" keyword - foreach disable_path: disabled_drivers - if drv_path == disable_path - build = false - reason = 'Explicitly disabled via build config' - endif - endforeach + if disabled_drivers.contains(drv_path) + build = false + reason = 'Explicitly disabled via build config' + else + # pull in driver directory which should update all the local variables + subdir(drv_path) + endif + if build # get dependency objs from strings shared_deps = ext_deps -- 2.17.1
> -----Original Message-----
> From: Bruce Richardson <bruce.richardson@intel.com>
> Sent: Monday, March 30, 2020 3:53 PM
> To: Stojaczyk, Dariusz <dariusz.stojaczyk@intel.com>
> Cc: dev@dpdk.org
> Subject: Re: [PATCH] build: don't parse build configs of explicitly disabled
> drivers
>
> [SNIP]
>
> Looking at this code and the meson docs again, I think this block and
> previous can be simplified by using the array "contains()" method to avoid
> the loop.
>
> if disabled_drivers.contains(drv_path)
> build = false
> reson = '....'
> else
> # pull in driver ...
> subdir(drv_path)
> endif
>
> Regards,
> /Bruce
Oops, I haven't seen that mail. Sorry! I just pushed v2.
D.
07/05/2020 12:13, Darek Stojaczyk:
> Even when a PMD was disabled with meson's disable_drivers option
> its config file was still being parsed. Some of the PMD configs
> attempt to find a library they depend on and parse its header files
> with certain assumptions. If the library is found, but it's simply
> too old to contain the necessary header files, the meson build
> fails and it can only be fixed by either updating that library, or
> expanding the meson script for the faulty PMD.
>
> While the latter should be still done for the sake of DPDK quality,
> an intermediate solution would be to skip building the faulty PMD
> - there's a chance we don't need it. That's what this patch allows.
>
> Signed-off-by: Darek Stojaczyk <dariusz.stojaczyk@intel.com>
> Acked-by: Bruce Richardson <bruce.richardson@intel.com>
> ---
> v2:
> - simplified the code with if X.contains(), as Bruce suggested
> - removed the related comment from the code - it should be
> self-documented now
Applied, thanks