DPDK patches and discussions
 help / color / mirror / Atom feed
From: "Juraj Linkeš" <juraj.linkes@pantheon.tech>
To: Bruce Richardson <bruce.richardson@intel.com>
Cc: "thomas@monjalon.net" <thomas@monjalon.net>,
	"Ruifeng.Wang@arm.com" <Ruifeng.Wang@arm.com>,
	"Honnappa.Nagarahalli@arm.com" <Honnappa.Nagarahalli@arm.com>,
	"jerinjacobk@gmail.com" <jerinjacobk@gmail.com>,
	"hemant.agrawal@nxp.com" <hemant.agrawal@nxp.com>,
	"ferruh.yigit@intel.com" <ferruh.yigit@intel.com>,
	"aboyer@pensando.io" <aboyer@pensando.io>,
	"dev@dpdk.org" <dev@dpdk.org>
Subject: Re: [dpdk-dev] [RFC PATCH v2] build: kni cross-compilation support
Date: Fri, 5 Feb 2021 09:26:05 +0000	[thread overview]
Message-ID: <0055021693274b0caf0b542bbefd0715@pantheon.tech> (raw)
In-Reply-To: <20210204173352.GE1712@bricha3-MOBL.ger.corp.intel.com>



> -----Original Message-----
> From: Bruce Richardson <bruce.richardson@intel.com>
> Sent: Thursday, February 4, 2021 6:34 PM
> To: Juraj Linkeš <juraj.linkes@pantheon.tech>
> Cc: thomas@monjalon.net; Ruifeng.Wang@arm.com;
> Honnappa.Nagarahalli@arm.com; jerinjacobk@gmail.com;
> hemant.agrawal@nxp.com; ferruh.yigit@intel.com; aboyer@pensando.io;
> dev@dpdk.org
> Subject: Re: [RFC PATCH v2] build: kni cross-compilation support
> 
> On Thu, Feb 04, 2021 at 10:51:41AM +0100, Juraj Linkeš wrote:
> > The kni linux module is using a custom target for building, which
> > doesn't take into account any cross compilation arguments. The
> > arguments in question are ARCH, CROSS_COMPILE (for gcc, clang) and CC,
> > LD (for clang). Get those from the cross file and pass them to the
> > custom target.
> >
> > The user supplied path may not contain the 'build' directory, such as
> > when using cross-compiled headers, so only append that in the default
> > case (when no path is supplied in native builds) and use the
> > unmodified path from the user otherwise. Also modify the install path
> accordingly.
> >
> > Signed-off-by: Juraj Linkeš <juraj.linkes@pantheon.tech>
> 
> Some comments inline below.
> 

Thanks, these are very helpful.

> /Bruce
> 
> > ---
> >  kernel/linux/kni/meson.build |  8 +++---
> >  kernel/linux/meson.build     | 56 ++++++++++++++++++++++++++++++++----
> >  2 files changed, 55 insertions(+), 9 deletions(-)
> >
> > diff --git a/kernel/linux/kni/meson.build
> > b/kernel/linux/kni/meson.build index 07e0c9dae7..bb7123354f 100644
> > --- a/kernel/linux/kni/meson.build
> > +++ b/kernel/linux/kni/meson.build
> > @@ -13,7 +13,7 @@ kni_sources = files(  custom_target('rte_kni',
> >  	input: kni_sources,
> >  	output: 'rte_kni.ko',
> > -	command: ['make', '-j4', '-C', kernel_dir + '/build',
> > +	command: ['make', '-j4', '-C', kernel_dir,
> >  		'M=' + meson.current_build_dir(),
> >  		'src=' + meson.current_source_dir(),
> >  		'MODULE_CFLAGS=-include ' + meson.source_root() +
> > '/config/rte_config.h' + @@ -21,8 +21,8 @@ custom_target('rte_kni',
> >  		' -I' + meson.source_root() + '/lib/librte_kni' +
> >  		' -I' + meson.build_root() +
> >  		' -I' + meson.current_source_dir(),
> > -		'modules'],
> > +		'modules'] + cross_args,
> >  	depends: kni_mkfile,
> > -	install: true,
> > -	install_dir: kernel_dir + '/extra/dpdk',
> > +	install: install,
> > +	install_dir: install_dir,
> >  	build_by_default: get_option('enable_kmods'))
> 
> rather than "kernel_dir" and "install_dir" can we rename these to
> "kernel_build_dir" and "kernel_install_dir" for clarity.
> 

Makes sense, I'll do that.

> > diff --git a/kernel/linux/meson.build b/kernel/linux/meson.build index
> > 5c864a4653..74097299bb 100644
> > --- a/kernel/linux/meson.build
> > +++ b/kernel/linux/meson.build
> > @@ -3,20 +3,66 @@
> >
> >  subdirs = ['kni']
> >
> > +kernel_version = run_command('uname', '-r').stdout().strip()
> 
> Rename to "host_kernel_version" and probably should be only queried in the
> native build case.
> 

In meson vernicular the host machine is where the binaries will be running, i.e. what we're building for, so this may be a bit confusing - we could name it build_machine_kernel_version.
In any case, I'll move to the the native build if branch. But then maybe we don't need to rename it?

> > +cross_args = []
> >  # if we are cross-compiling we need kernel_dir specified -if
> > get_option('kernel_dir') == '' and meson.is_cross_build()
> > -	error('Need "kernel_dir" option for kmod compilation when cross-
> compiling')
> > +if meson.is_cross_build()
> > +	if get_option('kernel_dir') == ''
> > +		error('Need "kernel_dir" option for kmod compilation when
> cross-compiling')
> > +	else
> > +		install_dir = ''
> > +		install = false
> 
> I think these should be defined and initialized further up the file, outside the
> conditional block.
> 

Ok, I'll make the default true and set it to false in the cross compilation case.

> > +		cross_compiler = find_program('c').path()
> > +		if cross_compiler.endswith('gcc')
> > +			cross_prefix = run_command([py3, '-c', 'print("' +
> cross_compiler + '"[:-3])']).stdout().strip()
> > +		elif cross_compiler.endswith('clang')
> > +			cross_prefix = ''
> > +			found_target = false
> > +			# search for '-target' and use the arg that follows
> > +			# (i.e. the value of '-target') as cross_prefix
> > +			foreach cross_c_arg :
> meson.get_cross_property('c_args')
> > +				if found_target and cross_prefix == ''
> > +					cross_prefix = cross_c_arg
> > +				endif
> > +				if cross_c_arg == '-target'
> > +					found_target = true
> > +				endif
> > +			endforeach
> > +			if cross_prefix == ''
> > +				error('Didn\'t find -target and its value in' +
> > +				      ' c_args in input cross-file.')
> > +			endif
> > +			linker = 'lld'
> > +			foreach cross_c_link_arg :
> meson.get_cross_property('c_link_args')
> > +				if cross_c_link_arg.startswith('-fuse-ld')
> > +					linker = cross_c_link_arg.split('=')[1]
> > +				endif
> > +			endforeach
> > +			cross_args += ['CC=@0@'.format(cross_compiler),
> 'LD=ld.@0@'.format(linker)]
> > +		else
> > +			error('Unsupported cross compiler:
> @0@'.format(cross_compiler))
> > +		endif
> > +		if host_machine.cpu_family() == 'aarch64'
> > +			cross_arch = 'arm64'
> > +		else
> > +			cross_arch = build_machine.cpu_family()
> > +		endif
> > +		cross_args += ['ARCH=@0@'.format(cross_arch),
> > +			'CROSS_COMPILE=@0@'.format(cross_prefix)]
> > +	endif
> > +else
> > +	install_dir = '/lib/modules/' + kernel_version + '/extra/dpdk'
> > +	install = true
> >  endif
> >
> 
> The block for cross-compiling is fairly large and complex, so I'm wondering how
> we can simplify things a bit. If we had multiple kernel modules I'd suggest
> splitting thing up into a native and cross-build subdirectories to get the build
> info, but that seems like overkill here.

This configuration would be the same for all kernel modules (right?), so I'm not sure how the number of kernel modules is relevant here.
If we split it, what would the dir structure look like? Something like this?
kernel/linux/
├── aarch64
├── native
├── kni
├── <other_mods>

> Instead I wonder if it would work better to
> handle all the native case initially in a hopefully simpler "if block" and then do
> subdir_done(), leaving everything else to be the cross-compilation recipe and
> saving at least one level of indentation.
> 

Wouldn't we need to duplicate the code that does make kernelversion and subdirs the actaul modules?

> Also, throughout the block, anywhere you have an "error()" call you can use
> "endif" instead of "else" and save more indentation space.
> 

Good point, I'll make the change.

> >  kernel_dir = get_option('kernel_dir')  if kernel_dir == ''
> >  	# use default path for native builds
> > -	kernel_version = run_command('uname', '-r').stdout().strip()
> > -	kernel_dir = '/lib/modules/' + kernel_version
> > +	kernel_dir = '/lib/modules/' + kernel_version + '/build'
> >  endif
> >
> >  # test running make in kernel directory, using "make kernelversion"
> > -make_returncode = run_command('make', '-sC', kernel_dir + '/build',
> > +make_returncode = run_command('make', '-sC', kernel_dir,
> >  		'kernelversion').returncode()
> >  if make_returncode != 0
> >  	error('Cannot compile kernel modules as requested - are kernel
> > headers installed?')
> > --
> > 2.20.1
> >


  reply	other threads:[~2021-02-05  9:26 UTC|newest]

Thread overview: 37+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2021-01-29 10:29 [dpdk-dev] [RFC PATCH v1] build: kni gcc " Juraj Linkeš
2021-01-29 11:43 ` Bruce Richardson
2021-01-29 12:33   ` Juraj Linkeš
2021-01-29 13:51     ` Bruce Richardson
2021-01-29 14:36       ` Juraj Linkeš
2021-01-29 14:42         ` Bruce Richardson
2021-01-29 14:47           ` Juraj Linkeš
2021-01-29 15:01             ` Bruce Richardson
2021-01-29 15:17               ` Juraj Linkeš
2021-01-29 15:39                 ` Bruce Richardson
2021-02-01  7:48                   ` Juraj Linkeš
2021-02-04  9:51 ` [dpdk-dev] [RFC PATCH v2] build: kni " Juraj Linkeš
2021-02-04 17:33   ` Bruce Richardson
2021-02-05  9:26     ` Juraj Linkeš [this message]
2021-02-05  9:38       ` Bruce Richardson
2021-02-05  9:44         ` Thomas Monjalon
2021-02-05  9:42       ` Bruce Richardson
2021-02-05 14:46   ` [dpdk-dev] [RFC PATCH v3] " Juraj Linkeš
2021-02-05 14:52     ` Bruce Richardson
2021-02-05 15:02       ` Juraj Linkeš
2021-02-05 15:04     ` [dpdk-dev] [RFC PATCH v4] " Juraj Linkeš
2021-02-05 15:27       ` Bruce Richardson
2021-02-08 10:17         ` Juraj Linkeš
2021-02-08 10:26           ` Bruce Richardson
2021-02-08 10:56             ` Thomas Monjalon
2021-02-08 11:05               ` Bruce Richardson
2021-02-08 11:21                 ` Thomas Monjalon
2021-02-08 11:45                   ` Bruce Richardson
2021-02-08 17:23                     ` Thomas Monjalon
2021-02-09  8:47       ` [dpdk-dev] [PATCH v5] " Juraj Linkeš
2021-02-09 11:50         ` Bruce Richardson
2021-02-09 12:07           ` Juraj Linkeš
2021-02-09 12:39             ` Bruce Richardson
2021-02-11 12:59         ` [dpdk-dev] [PATCH v6] " Juraj Linkeš
2021-03-09  8:47           ` Juraj Linkeš
2021-03-09 16:26             ` Andrew Boyer
2021-03-15 22:45           ` Thomas Monjalon

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=0055021693274b0caf0b542bbefd0715@pantheon.tech \
    --to=juraj.linkes@pantheon.tech \
    --cc=Honnappa.Nagarahalli@arm.com \
    --cc=Ruifeng.Wang@arm.com \
    --cc=aboyer@pensando.io \
    --cc=bruce.richardson@intel.com \
    --cc=dev@dpdk.org \
    --cc=ferruh.yigit@intel.com \
    --cc=hemant.agrawal@nxp.com \
    --cc=jerinjacobk@gmail.com \
    --cc=thomas@monjalon.net \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).