Skip to contentSkip to conversation

flashrom: bump to version 0.9.9

PATCH New Gary Bisson 2 participants 3 messages Apr 11, 2016

lore
Share & tools
lore
#0 lore raw

PATCH 1/1 flashrom: bump to version 0.9.9

Here are the release notes:
https://www.flashrom.org/Flashrom/0.9.9

Note that this version brings many cross-platform and cross-architecture improvements like musl/uclibc support.

Therefore the patches can be removed as now fixed upstream.

Also remove the architecture limitation since MIPS, PPC, ARM, PPC64, ARM64, SPARC, SH, M68K are all supported now.

---
Hi all,

I've had to use flashrom and realized the current package was limited to x86 architectures. Looking into it, it seems that most architectures are now supported so I simply removed the limitation.

However I haven't tried to build against all the possible architectures. Is it possible to cherry-pick (without merging) that patch and make the build server try them all?

One thing is for sure, it works great on ARM.

Regards,
Gary

---
 package/flashrom/0001-spi.patch      | 25 -------------------------
 package/flashrom/0002-sys-io.h.patch | 27 ---------------------------
 package/flashrom/Config.in           |  2 --
 package/flashrom/flashrom.hash       |  2 +-
 package/flashrom/flashrom.mk         |  2 +-
 5 files changed, 2 insertions(+), 56 deletions(-)
 delete mode 100644 package/flashrom/0001-spi.patch
 delete mode 100644 package/flashrom/0002-sys-io.h.patch
package/flashrom/0001-spi.patch deleted +0 -25
deleted file mode 100644
index 22ffcb5..0000000
@@ -1,25 +0,0 @@
-linux_spi: add missing include
-
-Some defines (e.g. _IOC_SIZEBITS) are defined in linux/ioctl.h,
-so it must be included before it is used, by SPI_IOC_MESSAGE
-from linux/spi/spidev.h
-
-Fixes build errors with the musl C library, as seen in these
-Buildroot autobuilder failures:
-
-    http://autobuild.buildroot.org/results/2a3/2a3744007c630c267575a638ebcd83a4b97644f5/build-end.log
-    http://autobuild.buildroot.org/results/3de/3de936d9be79e151e66af15193084d82a0f2c04a/build-end.log
-
-Signed-off-by: "Yann E. MORIN" <[email protected]>
-
-diff -durN flashrom-0.9.8.orig/linux_spi.c flashrom-0.9.8/linux_spi.c
---- flashrom-0.9.8.orig/linux_spi.c	2015-10-28 19:42:38.480285847 +0100
-+++ flashrom-0.9.8/linux_spi.c	2015-10-28 19:43:15.492994613 +0100
-@@ -27,6 +27,7 @@
- #include <ctype.h>
- #include <unistd.h>
- #include <linux/types.h>
-+#include <linux/ioctl.h>
- #include <linux/spi/spidev.h>
- #include <sys/ioctl.h>
- #include "flash.h"
package/flashrom/0002-sys-io.h.patch deleted +0 -27
deleted file mode 100644
index 8d990c6..0000000
@@ -1,27 +0,0 @@
-hwaccess: sys/io.h is not specific to glibc
-
-Under Linux, sys/io.h provides inb and outb, so we really need it.
-However, its inclusion is conditional to the _GLIBC_ define. This is
-usually OK under Linux, since both glibc and uClibc define it (uclibc
-fakes being glibc).
-
-But the musl C library does not impersonate glibc, so we're missing
-including sys/io.h in this case.
-
-Change the include from checking _GLIBC_ to checking whether this is
-Linux, looking for the __linux__ define.
-
-Signed-off-by: "Yann E. MORIN" <[email protected]>
-
-diff -durN flashrom-0.9.8.orig/hwaccess.h flashrom-0.9.8/hwaccess.h
---- flashrom-0.9.8.orig/hwaccess.h	2015-02-10 09:03:10.000000000 +0100
-+++ flashrom-0.9.8/hwaccess.h	2015-10-28 20:01:54.259202484 +0100
-@@ -27,7 +27,7 @@
- #include "platform.h"
- 
- #if IS_X86
--#if defined(__GLIBC__)
-+#if defined(__linux__)
- #include <sys/io.h>
- #endif
- #endif
package/flashrom/Config.in +0 -2
index adc068c..46fa641 100644
@@ -6,7 +6,6 @@ config BR2_PACKAGE_FLASHROM
 	select BR2_PACKAGE_LIBFTDI
 	# dmidecode is only a runtime dependency
 	select BR2_PACKAGE_DMIDECODE
-	depends on BR2_i386 || BR2_x86_64
 	depends on BR2_TOOLCHAIN_HAS_THREADS 	# libusb
 	help
 	  flashrom is a utility for identifying, reading, writing,
@@ -18,5 +17,4 @@ config BR2_PACKAGE_FLASHROM
 	  http://flashrom.org/
 
 comment "flashrom needs a toolchain w/ threads"
-	depends on BR2_i386 || BR2_x86_64
 	depends on !BR2_TOOLCHAIN_HAS_THREADS
package/flashrom/flashrom.hash +1 -1
index a216869..27328e4 100644
@@ -1,2 +1,2 @@
 # Locally computed
-sha256 13dc7c895e583111ecca370363a3527d237d178a134a94b20db7df177c05f934 flashrom-0.9.8.tar.bz2
+sha256 cb3156b0f63eb192024b76c0814135930297aac41f80761a5d293de769783c45 flashrom-0.9.9.tar.bz2
package/flashrom/flashrom.mk +1 -1
index 7e0fcf2..7ebda33 100644
@@ -4,7 +4,7 @@
 #
 ################################################################################
 
-FLASHROM_VERSION = 0.9.8
+FLASHROM_VERSION = 0.9.9
 FLASHROM_SOURCE = flashrom-$(FLASHROM_VERSION).tar.bz2
 FLASHROM_SITE = http://download.flashrom.org/releases
 FLASHROM_DEPENDENCIES = pciutils libusb libusb-compat libftdi host-pkgconf
Signature · 1 line
2.7.0
#1 lore raw

PATCH 1/1 flashrom: bump to version 0.9.9

Hello,

package/flashrom/Config.in config BR2_PACKAGE_FLASHROM · line 8

Quoted 8 lines · Gary Bisson
package/flashrom/Config.in +0 -0
index adc068c..46fa641 100644
@@ -6,7 +6,6 @@ config BR2_PACKAGE_FLASHROM
 	select BR2_PACKAGE_LIBFTDI
 	# dmidecode is only a runtime dependency
 	select BR2_PACKAGE_DMIDECODE

dmidecode is available on i386/x86-64 only, so you cannot keep this select and remove the i386/x86-64 dependency at the same time.

Thanks!

Thomas

Signature · 3 lines
Thomas Petazzoni, CTO, Free Electrons
Embedded Linux, Kernel and Android engineering
http://free-electrons.com
#2 lore raw

PATCH 1/1 flashrom: bump to version 0.9.9

Thomas, All,

Quoted 4 lines · Thomas Petazzoni

Hello,

On Mon, 11 Apr 2016 11:46:26 +0200, Gary Bisson wrote:

Quoted 11 lines

package/flashrom/Config.in config BR2_PACKAGE_FLASHROM · line 8

Quoted 8 lines
package/flashrom/Config.in +0 -0
index adc068c..46fa641 100644
@@ -6,7 +6,6 @@ config BR2_PACKAGE_FLASHROM
      select BR2_PACKAGE_LIBFTDI
      # dmidecode is only a runtime dependency
      select BR2_PACKAGE_DMIDECODE

dmidecode is available on i386/x86-64 only, so you cannot keep this select and remove the i386/x86-64 dependency at the same time.

You're right. Also, pci-utils is not available for bfin so I guess I need to exclude this architecture.

Regards,
Gary