From: Thomas Petazzoni via buildroot <buildroot@buildroot.org>
To: Alistair Francis <alistair23@gmail.com>
Cc: Samuel Martin <s.martin49@gmail.com>,
Alistair Francis <alistair.francis@wdc.com>,
buildroot@buildroot.org
Subject: Re: [Buildroot] [PATCH v2 2/2] package/libspdm: Initial commit
Date: Sun, 6 Aug 2023 15:56:06 +0200 [thread overview]
Message-ID: <20230806155606.53d6f567@windsurf> (raw)
In-Reply-To: <20230724014832.2207166-2-alistair.francis@wdc.com>
Hello,
On Mon, 24 Jul 2023 11:48:32 +1000
Alistair Francis <alistair23@gmail.com> wrote:
> Signed-off-by: Alistair Francis <alistair.francis@wdc.com>
Commit title should be:
package/libspdm: new package
> ---
> package/Config.in | 1 +
> package/libspdm/Config.in | 20 ++++++++++++
> package/libspdm/libspdm.hash | 3 ++
> package/libspdm/libspdm.mk | 59 ++++++++++++++++++++++++++++++++++++
> 4 files changed, 83 insertions(+)
> create mode 100644 package/libspdm/Config.in
> create mode 100644 package/libspdm/libspdm.hash
> create mode 100644 package/libspdm/libspdm.mk
A new entry in the DEVELOPERS file is needed.
> diff --git a/package/libspdm/Config.in b/package/libspdm/Config.in
> new file mode 100644
> index 0000000000..b732662842
> --- /dev/null
> +++ b/package/libspdm/Config.in
> @@ -0,0 +1,20 @@
> +config BR2_PACKAGE_LIBSPDM
> + bool "libspdm"
> + select BR2_PACKAGE_OPENSSL
> + select BR2_PACKAGE_OPENSSL_FORCE_LIBOPENSSL
> + help
> + libspdm is a sample implementation that follows
> + the DMTF SPDM specifications
> +
> + https://github.com/DMTF/libspdm
> +
> +config BR2_PACKAGE_LIBSPDM_CPU_FAMILLY
> + string
> + default "arc" if BR2_arcle || BR2_arceb
> + default "arm" if BR2_arm || BR2_armeb
> + default "aarch64" if BR2_aarch64 || BR2_aarch64_be
> + default "ia32" if BR2_i386
> + default "riscv32" if BR2_riscv && BR2_RISCV_32
> + default "riscv64" if BR2_riscv && BR2_RISCV_64
> + default "x64" if BR2_x86_64
> + default BR2_ARCH # Yes, this should work...\
Are all CPU families really supported? Usually, when a package needs to
know the CPU family/platform, it is a very strong sign that the package
has some architecture-specific stuff, and that some of our "special"
architectures will not be supported.
According to the CMakeLists.txt, I'm pretty correct:
if(ARCH STREQUAL "x64")
MESSAGE("ARCH = x64")
elseif(ARCH STREQUAL "ia32")
MESSAGE("ARCH = ia32")
elseif(ARCH STREQUAL "arm")
MESSAGE("ARCH = arm")
elseif(ARCH STREQUAL "aarch64")
MESSAGE("ARCH = aarch64")
elseif(ARCH STREQUAL "riscv32")
MESSAGE("ARCH = riscv32")
elseif(ARCH STREQUAL "riscv64")
MESSAGE("ARCH = riscv64")
elseif(ARCH STREQUAL "arc")
MESSAGE("ARCH = arc")
elseif(ARCH STREQUAL "nios2")
MESSAGE("ARCH = nios2")
else()
MESSAGE(FATAL_ERROR "Unknown ARCH")
endif()
Your package will fail to build on any other architectures than the
ones listed here.
> diff --git a/package/libspdm/libspdm.mk b/package/libspdm/libspdm.mk
> new file mode 100644
> index 0000000000..daf6639808
> --- /dev/null
> +++ b/package/libspdm/libspdm.mk
> @@ -0,0 +1,59 @@
> +################################################################################
> +#
> +# libspdm
> +#
> +################################################################################
> +
> +LIBSPDM_VERSION = 3.0.0
> +LIBSPDM_SITE = $(call github,DMTF,libspdm,$(LIBSPDM_VERSION))
> +LIBSPDM_LICENSE = BSD-3-Clause
> +LIBSPDM_LICENSE_FILES = LICENSE.md
> +
> +LIBSPDM_INSTALL_STAGING = YES
> +LIBSPDM_INSTALL_TARGET = YES
This line is not needed, it's the default.
> +
> +LIBSPDM_DEPENDENCIES += openssl
Just =, not +=.
> +
> +LIBSPDM_TARGET_CPU_FAMILY = $(call qstrip,$(BR2_PACKAGE_LIBSPDM_CPU_FAMILLY))
> +
> +LIBSPDM_CONF_OPTS = \
> + -DARCH=$(LIBSPDM_TARGET_CPU_FAMILY) \
> + -DTOOLCHAIN=NONE \
> + -DTARGET=Release \
> + -DCRYPTO=openssl \
> + -DENABLE_BINARY_BUILD=1 \
> + -DCOMPILED_LIBCRYPTO_PATH=/usr/lib/ \
> + -DCOMPILED_LIBSSL_PATH=/usr/lib/ \
> + -DDISABLE_TESTS=1 \
> + -DLINK_FLAGS=$(STAGING_DIR)
> +
> +define LIBSPDM_INSTALL_STAGING_CMDS
> + $(INSTALL) -m 0755 -t $(STAGING_DIR)/usr/lib/ $(@D)/lib/*
> +
> + $(INSTALL) -d $(STAGING_DIR)/usr/include/libspdm/library
> + $(INSTALL) -D -m 0755 $(@D)/include/library/* $(STAGING_DIR)/usr/include/libspdm/library
> +
> + $(INSTALL) -d $(STAGING_DIR)/usr/include/libspdm/internal
> + $(INSTALL) -D -m 0755 $(@D)/include/internal/*.h $(STAGING_DIR)/usr/include/libspdm/internal/
> +
> + $(INSTALL) -d $(STAGING_DIR)/usr/include/libspdm/hal/library/cryptlib
> + $(INSTALL) -d $(STAGING_DIR)/usr/include/libspdm/hal/library/requester
> + $(INSTALL) -d $(STAGING_DIR)/usr/include/libspdm/hal/library/responder
> + $(INSTALL) -D -m 0755 $(@D)/include/hal/*.h $(STAGING_DIR)/usr/include/libspdm/hal/
> + $(INSTALL) -D -m 0755 $(@D)/include/hal/library/*.h $(STAGING_DIR)/usr/include/libspdm/hal/library/
> + $(INSTALL) -D -m 0755 $(@D)/include/hal/library/cryptlib/* $(STAGING_DIR)/usr/include/libspdm/hal/library/cryptlib/
> + $(INSTALL) -D -m 0755 $(@D)/include/hal/library/requester/* $(STAGING_DIR)/usr/include/libspdm/hal/library/requester/
> + $(INSTALL) -D -m 0755 $(@D)/include/hal/library/responder/* $(STAGING_DIR)/usr/include/libspdm/hal/library/responder/
> +
> + $(INSTALL) -d $(STAGING_DIR)/usr/include/libspdm/industry_standard
> + $(INSTALL) -D -m 0755 $(@D)/include/industry_standard/* $(STAGING_DIR)/usr/include/libspdm/industry_standard
> +
> + $(INSTALL) -d $(STAGING_DIR)/usr/include/libspdm/os_stub/spdm_crypt_ext_lib
> + $(INSTALL) -D -m 0755 $(@D)/os_stub/spdm_crypt_ext_lib/*.h $(STAGING_DIR)/usr/include/libspdm/os_stub/spdm_crypt_ext_lib
This seems very verbose. If you're copying all files/directories of
$(@D)/include/* to $(STAGING_DIR)/usr/include/libspdm, why not:
mkdir -p $(STAGING_DIR)/usr/include/libspdm/
cp -dpfr $(@D)/include/* $(STAGING_DIR)/usr/include/libspdm/
> +endef
> +
> +define LIBSPDM_INSTALL_TARGET_CMDS
> + $(INSTALL) -D -m 0755 $(@D)/lib/* $(TARGET_DIR)/usr/lib/
This is not good, as it could create a file called "lib" in
$(TARGET_DIR)/usr if $(TARGET_DIR)/usr/lib doesn't already exist. So:
mkdir -p $(TARGET_DIR)/usr/lib
cp -dpfr $(@D)/lib/* $(TARGET_DIR)/usr/lib/
Thanks,
Thomas
--
Thomas Petazzoni, co-owner and CEO, Bootlin
Embedded Linux and Kernel engineering and training
https://bootlin.com
_______________________________________________
buildroot mailing list
buildroot@buildroot.org
https://lists.buildroot.org/mailman/listinfo/buildroot
next prev parent reply other threads:[~2023-08-06 13:57 UTC|newest]
Thread overview: 5+ messages / expand[flat|nested] mbox.gz Atom feed top
2023-07-24 1:48 [Buildroot] [PATCH v2 1/2] package/libopenssl: Copy the crypto include files Alistair Francis
2023-07-24 1:48 ` [Buildroot] [PATCH v2 2/2] package/libspdm: Initial commit Alistair Francis
2023-08-06 13:56 ` Thomas Petazzoni via buildroot [this message]
2023-08-06 13:49 ` [Buildroot] [PATCH v2 1/2] package/libopenssl: Copy the crypto include files Thomas Petazzoni via buildroot
2023-08-11 20:28 ` Alistair Francis
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=20230806155606.53d6f567@windsurf \
--to=buildroot@buildroot.org \
--cc=alistair.francis@wdc.com \
--cc=alistair23@gmail.com \
--cc=s.martin49@gmail.com \
--cc=thomas.petazzoni@bootlin.com \
/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