Skip to contentSkip to conversation

lua

PATCH Superseded Francois Perrad 4 participants 19 messages Mar 1, 2014 – Jul 18, 2014

v3 of 4 Superseded Latest is v4 →

lore
Share & tools
lore

Francois Perrad

author
#0 lore raw

v3 cover lua

delta V3 vs V1 (remaining)
- luajit: refactor without dont-strip.patch now with 'true' instead of '@:'
- luajit: handles BR2_PREFER_STATIC_LIB
new patch
- luainterpreter: add host variant
new patch, previously included in the luarocks one
- ljlinenoise
split dependencies for comment statement in Config.in
- lua-iconv
add missing dependencies in .mk

delta V2 vs V1
- only a subset for 2014.02
- add fix for Blackfin FLAT

Francois Perrad (8):
luajit: refactor without dont-strip.patch
luajit: refactor without shared-lib.patch
luajit: handles BR2_PREFER_STATIC_LIB
luajit: add host variant
luainterpreter: add host variant
luarocks: allow to work with host-luainterpreter
ljlinenoise: new package
lua-iconv: new package

 package/Config.in                         |    2 ++
 package/ljlinenoise/Config.in             |   16 ++++++++++++++++
 package/ljlinenoise/ljlinenoise.mk        |   13 +++++++++++++
 package/lua-iconv/Config.in               |    8 ++++++++
 package/lua-iconv/lua-iconv.mk            |   14 ++++++++++++++
 package/luainterpreter/luainterpreter.mk  |    4 ++++
 package/luajit/luajit-02-shared-lib.patch |   17 -----------------
 package/luajit/luajit-03-dont-strip.patch |   23 -----------------------
 package/luajit/luajit.mk                  |   18 +++++++++++++++++-
 package/luarocks/luarocks.mk              |   18 ++++++++++++------
 10 files changed, 86 insertions(+), 47 deletions(-)
 create mode 100644 package/ljlinenoise/Config.in
 create mode 100644 package/ljlinenoise/ljlinenoise.mk
 create mode 100644 package/lua-iconv/Config.in
 create mode 100644 package/lua-iconv/lua-iconv.mk
 delete mode 100644 package/luajit/luajit-02-shared-lib.patch
 delete mode 100644 package/luajit/luajit-03-dont-strip.patch
Signature · 1 line
1.7.9.5
author
#1 lore raw

PATCH v3 1/8 luajit: refactor without dont-strip.patch

the variable TARGET_STRIP does the job

Signed-off-by Francois Perrad <[email protected]>
---
 package/luajit/luajit-03-dont-strip.patch |   23 -----------------------
 package/luajit/luajit.mk                  |    2 +-
 2 files changed, 1 insertion(+), 24 deletions(-)
 delete mode 100644 package/luajit/luajit-03-dont-strip.patch
package/luajit/luajit-03-dont-strip.patch deleted +0 -23
deleted file mode 100644
index 66bdadf..0000000
@@ -1,23 +0,0 @@
-Don't strip the library and binary
-
-This allows Buildroot to do it if needed, but only if needed.
-
-Signed-off-by: Thomas Petazzoni <[email protected]>
-
-Index: b/src/Makefile
-===================================================================
---- a/src/Makefile
-+++ b/src/Makefile
-@@ -665,12 +665,10 @@
- $(LUAJIT_SO): $(LJVMCORE_O)
- 	$(E) "DYNLINK   $@"
- 	$(Q)$(TARGET_LD) $(TARGET_ASHLDFLAGS) -o $@ $(LJVMCORE_DYNO) $(TARGET_ALIBS)
--	$(Q)$(TARGET_STRIP) $@
- 
- $(LUAJIT_T): $(TARGET_O) $(LUAJIT_O) $(TARGET_DEP)
- 	$(E) "LINK      $@"
- 	$(Q)$(TARGET_LD) $(TARGET_ALDFLAGS) -o $@ $(LUAJIT_O) -lluajit -L. $(TARGET_ALIBS)
--	$(Q)$(TARGET_STRIP) $@
- 	$(E) "OK        Successfully built LuaJIT"
- 
- ##############################################################################
package/luajit/luajit.mk +1 -1
index 4122d0a..72720d8 100644
@@ -37,7 +37,7 @@ define LUAJIT_BUILD_CMDS
 		DYNAMIC_CC="$(TARGET_CC) -fPIC" \
 		TARGET_LD="$(TARGET_CC)" \
 		TARGET_AR="$(TARGET_AR) rcus" \
-		TARGET_STRIP="$(TARGET_STRIP)" \
+		TARGET_STRIP=true \
 		TARGET_CFLAGS="$(TARGET_CFLAGS)" \
 		TARGET_LDFLAGS="$(TARGET_LDFLAGS)" \
 		HOST_CC="$(LUAJIT_HOST_CC)" \
Signature · 1 line
1.7.9.5
author
#2 lore raw

PATCH v3 2/8 luajit: refactor without shared-lib.patch

In reply to Francois Perrad

the variable BUILDMODE does the job

Signed-off-by Francois Perrad <[email protected]>
---
 package/luajit/luajit-02-shared-lib.patch |   17 -----------------
 package/luajit/luajit.mk                  |    1 +
 2 files changed, 1 insertion(+), 17 deletions(-)
 delete mode 100644 package/luajit/luajit-02-shared-lib.patch
package/luajit/luajit-02-shared-lib.patch deleted +0 -17
deleted file mode 100644
index 361e5c2..0000000
@@ -1,17 +0,0 @@
-Compile the luajit binary dynamically against the luajit library
-
-Signed-off-by: Thomas Petazzoni <[email protected]>
-
-Index: b/src/Makefile
-===================================================================
---- a/src/Makefile
-+++ b/src/Makefile
-@@ -669,7 +669,7 @@
- 
- $(LUAJIT_T): $(TARGET_O) $(LUAJIT_O) $(TARGET_DEP)
- 	$(E) "LINK      $@"
--	$(Q)$(TARGET_LD) $(TARGET_ALDFLAGS) -o $@ $(LUAJIT_O) $(TARGET_O) $(TARGET_ALIBS)
-+	$(Q)$(TARGET_LD) $(TARGET_ALDFLAGS) -o $@ $(LUAJIT_O) -lluajit -L. $(TARGET_ALIBS)
- 	$(Q)$(TARGET_STRIP) $@
- 	$(E) "OK        Successfully built LuaJIT"
-
package/luajit/luajit.mk +1 -0
index 72720d8..32596a1 100644
@@ -44,6 +44,7 @@ define LUAJIT_BUILD_CMDS
 		HOST_CFLAGS="$(HOST_CFLAGS)" \
 		HOST_LDFLAGS="$(HOST_LDFLAGS)" \
 		$(LUAJIT_NO_LARGEFILE) \
+		BUILDMODE=dynamic \
 		-C $(@D) amalg
 endef
Signature · 1 line
1.7.9.5
author
#3 lore raw

PATCH v3 3/8 luajit: handles BR2_PREFER_STATIC_LIB

In reply to Francois Perrad
Signed-off-by Francois Perrad <[email protected]>
---
 package/luajit/luajit.mk |    8 +++++++-
 1 file changed, 7 insertions(+), 1 deletion(-)
package/luajit/luajit.mk +7 -1
index 32596a1..859425a 100644
@@ -16,6 +16,12 @@ ifneq ($(BR2_LARGEFILE),y)
 LUAJIT_NO_LARGEFILE = TARGET_LFSFLAGS=
 endif
 
+ifeq ($(BR2_PREFER_STATIC_LIB),y)
+LUAJIT_BUILDMODE = static
+else
+LUAJIT_BUILDMODE = dynamic
+endif
+
 # The luajit build procedure requires the host compiler to have the
 # same bitness as the target compiler. Therefore, on a x86 build
 # machine, we can't build luajit for x86_64, which is checked in

@@ -44,7 +50,7 @@ define LUAJIT_BUILD_CMDS
 		HOST_CFLAGS="$(HOST_CFLAGS)" \
 		HOST_LDFLAGS="$(HOST_LDFLAGS)" \
 		$(LUAJIT_NO_LARGEFILE) \
-		BUILDMODE=dynamic \
+		BUILDMODE=$(LUAJIT_BUILDMODE) \
 		-C $(@D) amalg
 endef
Signature · 1 line
1.7.9.5
author
#4 lore raw

PATCH v3 4/8 luajit: add host variant

In reply to Francois Perrad
Signed-off-by Francois Perrad <[email protected]>
---
 package/luajit/luajit.mk |    9 +++++++++
 1 file changed, 9 insertions(+)
package/luajit/luajit.mk +9 -0
index 859425a..cb1d94e 100644
@@ -62,4 +62,13 @@ define LUAJIT_INSTALL_TARGET_CMDS
 	$(MAKE) PREFIX="/usr" DESTDIR="$(TARGET_DIR)" LDCONFIG=true -C $(@D) install
 endef
 
+define HOST_LUAJIT_BUILD_CMDS
+	$(MAKE) PREFIX="/usr" BUILDMODE=static -C $(@D) amalg
+endef
+
+define HOST_LUAJIT_INSTALL_CMDS
+	$(MAKE) PREFIX="/usr" DESTDIR="$(HOST_DIR)" -C $(@D) install
+endef
+
 $(eval $(generic-package))
+$(eval $(host-generic-package))
Signature · 1 line
1.7.9.5
#5 lore raw

PATCH v3 4/8 luajit: add host variant

Fran?ois, All,

On 2014-03-01 18:23 +0100, Francois Perrad spake thusly:

Quoted 14 lines
Signed-off-by Francois Perrad <[email protected]>
---
 package/luajit/luajit.mk |    9 +++++++++
 1 file changed, 9 insertions(+)
package/luajit/luajit.mk +0 -0
index 859425a..cb1d94e 100644
@@ -62,4 +62,13 @@ define LUAJIT_INSTALL_TARGET_CMDS
 	$(MAKE) PREFIX="/usr" DESTDIR="$(TARGET_DIR)" LDCONFIG=true -C $(@D) install
 endef

package/luajit/luajit.mk define LUAJIT_INSTALL_TARGET_CMDS · line 74

Quoted 10 lines
package/luajit/luajit.mk +9 -0
+define HOST_LUAJIT_BUILD_CMDS
+	$(MAKE) PREFIX="/usr" BUILDMODE=static -C $(@D) amalg
+endef
+
+define HOST_LUAJIT_INSTALL_CMDS
+	$(MAKE) PREFIX="/usr" DESTDIR="$(HOST_DIR)" -C $(@D) install
+endef
+
 $(eval $(generic-package))
+$(eval $(host-generic-package))

I'm really worried about how this is gonna play with the virtual-package insfrastructure curently pending.

On one hand, I'd like the virtual-package infra to go in first (hey! I wrote it! ;-) ), since I'm not very versed in this lua stuff (but you are), and we currently have no host virtual packages. But that would put the burden onto you to rebase your series.

On the other hand, sicne you're introducing host virtual packages, we may want to have them in before the virtual-package infra, so we get a pretty good idea of what is actually required.

So, whatever... Let's see what Thomas (deputy-maintainer while Peter's on vacation) will say about this. Thomas? ;-)

Regards,
Yann E. MORIN.

Signature · 6 lines
.-----------------.--------------------.------------------.--------------------.
|  Yann E. MORIN  | Real-Time Embedded | /"\ ASCII RIBBON | Erics' conspiracy: |
| +33 662 376 056 | Software  Designer | \ / CAMPAIGN     |  ___               |
| +33 223 225 172 `------------.-------:  X  AGAINST      |  \e/  There is no  |
| http://ymorin.is-a-geek.org/ | _/*\_ | / \ HTML MAIL    |   v   conspiracy.  |
'------------------------------^-------^------------------^--------------------'
#6 lore raw

PATCH v3 4/8 luajit: add host variant

Dear Yann E. MORIN,

Quoted 8 lines · Yann E. MORIN

I'm really worried about how this is gonna play with the virtual-package insfrastructure curently pending.

On one hand, I'd like the virtual-package infra to go in first (hey! I wrote it! ;-) ), since I'm not very versed in this lua stuff (but you are), and we currently have no host virtual packages. But that would put the burden onto you to rebase your series.

Quoted 6 lines

On the other hand, sicne you're introducing host virtual packages, we may want to have them in before the virtual-package infra, so we get a pretty good idea of what is actually required.

So, whatever... Let's see what Thomas (deputy-maintainer while Peter's on vacation) will say about this. Thomas? ;-)

While in general I prefer to merge things as soon as possible, and then improve on top of it later on, in this case, I believe it would be nice if the virtual package infra could from the beginning support the host package case since we know we will need it for the Lua stuff.

Yann, you could just test it with some simple dummy host packages to ensure that your virtual infra works for host packages, and leave it to Fran?ois to rebase his series on top of the infra once it has been merged.

How does it sounds?

Thomas

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

PATCH v3 4/8 luajit: add host variant

Thomas, All,

On 2014-03-01 19:39 +0100, Thomas Petazzoni spake thusly:

Quoted 3 lines · Yann E. MORIN
Quoted 2 lines

I'm really worried about how this is gonna play with the virtual-package insfrastructure curently pending.

[--SNIP--]

Quoted 11 lines

While in general I prefer to merge things as soon as possible, and then improve on top of it later on, in this case, I believe it would be nice if the virtual package infra could from the beginning support the host package case since we know we will need it for the Lua stuff.

Yann, you could just test it with some simple dummy host packages to ensure that your virtual infra works for host packages, and leave it to Fran?ois to rebase his series on top of the infra once it has been merged.

How does it sounds?

I already thought about this. But writing a host virtual package that actually makes sense is not so easy. But I can try to do so...

Regards,
Yann E. MORIN.

Signature · 6 lines
.-----------------.--------------------.------------------.--------------------.
|  Yann E. MORIN  | Real-Time Embedded | /"\ ASCII RIBBON | Erics' conspiracy: |
| +33 662 376 056 | Software  Designer | \ / CAMPAIGN     |  ___               |
| +33 223 225 172 `------------.-------:  X  AGAINST      |  \e/  There is no  |
| http://ymorin.is-a-geek.org/ | _/*\_ | / \ HTML MAIL    |   v   conspiracy.  |
'------------------------------^-------^------------------^--------------------'
author
#8 lore raw

PATCH v3 5/8 luainterpreter: add host variant

In reply to Francois Perrad
Signed-off-by Francois Perrad <[email protected]>
---
 package/luainterpreter/luainterpreter.mk |    2 ++
 1 file changed, 2 insertions(+)
package/luainterpreter/luainterpreter.mk +2 -0
index c37d621..976fec8 100644
@@ -5,8 +5,10 @@
 #############################################################
 
 LUAINTERPRETER_SOURCE =
+HOST_LUAINTERPRETER_SOURCE =
 LUAINTERPRETER_DEPENDENCIES = $(call qstrip,$(BR2_PACKAGE_PROVIDES_LUA_INTERPRETER))
 
 LUAINTERPRETER_ABIVER = $(call qstrip,$(BR2_PACKAGE_LUAINTERPRETER_ABI_VERSION))
 
 $(eval $(generic-package))
+$(eval $(host-generic-package))
Signature · 1 line
1.7.9.5
author
#9 lore raw

PATCH v3 6/8 luarocks: allow to work with host-luainterpreter

In reply to Francois Perrad

when we embed luajit on target, we use luarocks with host-luajit (instead of host-lua)

Signed-off-by Francois Perrad <[email protected]>
---
 package/luainterpreter/luainterpreter.mk |    2 ++
 package/luarocks/luarocks.mk             |   18 ++++++++++++------
 2 files changed, 14 insertions(+), 6 deletions(-)
package/luainterpreter/luainterpreter.mk +2 -0
index 976fec8..d5b95db 100644
@@ -12,3 +12,5 @@ LUAINTERPRETER_ABIVER = $(call qstrip,$(BR2_PACKAGE_LUAINTERPRETER_ABI_VERSION))
 
 $(eval $(generic-package))
 $(eval $(host-generic-package))
+
+LUA_RUN = $(HOST_DIR)/usr/bin/$(call qstrip,$(BR2_PACKAGE_PROVIDES_LUA_INTERPRETER))
package/luarocks/luarocks.mk +12 -6
index 86e57d2..0b99de6 100644
@@ -9,16 +9,22 @@ LUAROCKS_SITE = http://luarocks.org/releases/
 LUAROCKS_LICENSE = MIT
 LUAROCKS_LICENSE_FILES = COPYING
 
-HOST_LUAROCKS_DEPENDENCIES = host-lua luainterpreter
+HOST_LUAROCKS_DEPENDENCIES = host-luainterpreter luainterpreter
 
 LUAROCKS_CONFIG_DIR  = $(HOST_DIR)/usr/etc/luarocks
 LUAROCKS_CONFIG_FILE = $(LUAROCKS_CONFIG_DIR)/config-$(LUAINTERPRETER_ABIVER).lua
 
+HOST_LUAROCKS_CONF_OPT = \
+	--prefix=$(HOST_DIR)/usr \
+	--sysconfdir=$(LUAROCKS_CONFIG_DIR) \
+	--with-lua=$(HOST_DIR)/usr
+
+ifeq ($(BR2_PACKAGE_LUAJIT),y)
+HOST_LUAROCKS_CONF_OPT += --lua-suffix=jit
+endif
+
 define HOST_LUAROCKS_CONFIGURE_CMDS
-	cd $(@D) && ./configure \
-		--prefix=$(HOST_DIR)/usr \
-		--sysconfdir=$(LUAROCKS_CONFIG_DIR) \
-		--with-lua=$(HOST_DIR)/usr
+	cd $(@D) && ./configure $(HOST_LUAROCKS_CONF_OPT)
 endef
 
 define HOST_LUAROCKS_INSTALL_CMDS

@@ -43,4 +49,4 @@ endef
 $(eval $(host-generic-package))
 
 LUAROCKS_RUN = LUA_PATH="$(HOST_DIR)/usr/share/lua/$(LUAINTERPRETER_ABIVER)/?.lua" \
-	$(HOST_DIR)/usr/bin/lua $(HOST_DIR)/usr/bin/luarocks
+	$(LUA_RUN) $(HOST_DIR)/usr/bin/luarocks
Signature · 1 line
1.7.9.5
author
#10 lore raw

PATCH v3 7/8 ljlinenoise: new package

In reply to Francois Perrad
Signed-off-by Francois Perrad <[email protected]>
---
 package/Config.in                  |    1 +
 package/ljlinenoise/Config.in      |   16 ++++++++++++++++
 package/ljlinenoise/ljlinenoise.mk |   13 +++++++++++++
 3 files changed, 30 insertions(+)
 create mode 100644 package/ljlinenoise/Config.in
 create mode 100644 package/ljlinenoise/ljlinenoise.mk
package/Config.in +1 -0
index 46c4ba4..4a91f48 100644
@@ -369,6 +369,7 @@ source "package/cgilua/Config.in"
 source "package/copas/Config.in"
 source "package/coxpcall/Config.in"
 source "package/lbase64/Config.in"
+source "package/ljlinenoise/Config.in"
 source "package/ljsyscall/Config.in"
 source "package/lua-cjson/Config.in"
 source "package/lua-ev/Config.in"
package/ljlinenoise/Config.in new file +16 -0
new file mode 100644
index 0000000..4e0116e
@@ -0,0 +1,16 @@
+config BR2_PACKAGE_LJLINENOISE
+	bool "ljlinenoise"
+	select BR2_PACKAGE_LJSYSCALL
+	# ljsyscall is specifically for LuaJIT, not Lua.
+	depends on BR2_PACKAGE_LUAJIT
+	# ljsyscall is only available for some target architectures
+	depends on BR2_i386 || BR2_x86_64 || BR2_powerpc || BR2_arm || BR2_armeb
+	help
+	  ljlinenoise is a pure LuaJIT port of linenoise,
+	  a small alternative to readline and libedit.
+
+	  http://fperrad.github.io/ljlinenoise/
+
+comment "ljlinenoise needs LuaJIT"
+	depends on !BR2_PACKAGE_LUAJIT
+	depends on BR2_i386 || BR2_x86_64 || BR2_powerpc || BR2_arm || BR2_armeb
package/ljlinenoise/ljlinenoise.mk new file +13 -0
new file mode 100644
index 0000000..69f6130
@@ -0,0 +1,13 @@
+################################################################################
+#
+# ljlinenoise
+#
+################################################################################
+
+LJLINENOISE_VERSION_UPSTREAM = 0.1.1
+LJLINENOISE_VERSION = $(LJLINENOISE_VERSION_UPSTREAM)-1
+LJLINENOISE_SUBDIR  = ljlinenoise-$(LJLINENOISE_VERSION_UPSTREAM)
+LJLINENOISE_LICENSE = MIT
+LJLINENOISE_LICENSE_FILES = $(LJLINENOISE_SUBDIR)/COPYRIGHT
+
+$(eval $(luarocks-package))
Signature · 1 line
1.7.9.5
Applied
#11 lore raw

PATCH v3 7/8 ljlinenoise: new package

Dear Francois Perrad,

Quoted 9 lines · Francois Perrad
Signed-off-by Francois Perrad <[email protected]>
---
 package/Config.in                  |    1 +
 package/ljlinenoise/Config.in      |   16 ++++++++++++++++
 package/ljlinenoise/ljlinenoise.mk |   13 +++++++++++++
 3 files changed, 30 insertions(+)
 create mode 100644 package/ljlinenoise/Config.in
 create mode 100644 package/ljlinenoise/ljlinenoise.mk

Applied, after adjusting the Config.in file to indicate that ljsyscall is a runtime dependency only.

Thanks!

Thomas

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

PATCH v3 8/8 lua-iconv: new package

In reply to Francois Perrad
Signed-off-by Francois Perrad <[email protected]>
---
 package/Config.in              |    1 +
 package/lua-iconv/Config.in    |    8 ++++++++
 package/lua-iconv/lua-iconv.mk |   14 ++++++++++++++
 3 files changed, 23 insertions(+)
 create mode 100644 package/lua-iconv/Config.in
 create mode 100644 package/lua-iconv/lua-iconv.mk
package/Config.in +1 -0
index 4a91f48..3558106 100644
@@ -373,6 +373,7 @@ source "package/ljlinenoise/Config.in"
 source "package/ljsyscall/Config.in"
 source "package/lua-cjson/Config.in"
 source "package/lua-ev/Config.in"
+source "package/lua-iconv/Config.in"
 source "package/lua-msgpack-native/Config.in"
 source "package/luasql-sqlite3/Config.in"
 source "package/luabitop/Config.in"
package/lua-iconv/Config.in new file +8 -0
new file mode 100644
index 0000000..cad7c7c
@@ -0,0 +1,8 @@
+config BR2_PACKAGE_LUA_ICONV
+	bool "lua-iconv"
+	select BR2_PACKAGE_LIBICONV
+	depends on !BR2_ENABLE_LOCALE
+	help
+	  Lua binding to the POSIX 'iconv' library
+
+	  http://ittner.github.io/lua-iconv/
package/lua-iconv/lua-iconv.mk new file +14 -0
new file mode 100644
index 0000000..22f66de
@@ -0,0 +1,14 @@
+################################################################################
+#
+# lua-iconv
+#
+################################################################################
+
+LUA_ICONV_VERSION_UPSTREAM = 7
+LUA_ICONV_VERSION = $(LUA_ICONV_VERSION_UPSTREAM)-1
+LUA_ICONV_SUBDIR  = lua-iconv-$(LUA_ICONV_VERSION_UPSTREAM)
+LUA_ICONV_DEPENDENCIES = luainterpreter libiconv
+LUA_ICONV_LICENSE = MIT
+LUA_ICONV_LICENSE_FILES = $(LUA_ICONV_SUBDIR)/COPYING
+
+$(eval $(luarocks-package))
Signature · 1 line
1.7.9.5
#13 lore raw

v3 cover lua

In reply to Francois Perrad

Dear Francois Perrad,

Quoted 7 lines · Francois Perrad

Francois Perrad (8):
luajit: refactor without dont-strip.patch
luajit: refactor without shared-lib.patch
luajit: handles BR2_PREFER_STATIC_LIB
luajit: add host variant
luainterpreter: add host variant
luarocks: allow to work with host-luainterpreter

For all these patches, and especially the last three ones, I want commit logs. I really do appreciate _a lot_ all the work you're doing on Lua and Perl, it's really awesome. And I really mean it. But you need to work on improving your commit logs. For now, we have no idea *why* you're doing these changes.

All the other core developers are doing efforts to write long, descriptive commit logs, and experience has shown that they are immensely useful 6 or 12 months later when we got back in the Git history to understand why things were done. Look at what they do, and follow the basic rule of thumb that any non-trivial patch should have at least 2 or 3 paragraphs in its commit log explaining *why* the change is being made.

At this point, I have no idea what you are doing. I presume patches 4 and 5 are preparations for patch 6, but I don't understand what patch 6 is doing, and nowhere it is mentioned why patches 4 and 5 are needed.

Thanks!

Thomas

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

v3 cover lua

Fran?ois,

Quoted 22 lines · Thomas Petazzoni
Quoted 7 lines

Francois Perrad (8):
luajit: refactor without dont-strip.patch
luajit: refactor without shared-lib.patch
luajit: handles BR2_PREFER_STATIC_LIB
luajit: add host variant
luainterpreter: add host variant
luarocks: allow to work with host-luainterpreter

For all these patches, and especially the last three ones, I want commit logs. I really do appreciate _a lot_ all the work you're doing on Lua and Perl, it's really awesome. And I really mean it. But you need to work on improving your commit logs. For now, we have no idea *why* you're doing these changes.

All the other core developers are doing efforts to write long, descriptive commit logs, and experience has shown that they are immensely useful 6 or 12 months later when we got back in the Git history to understand why things were done. Look at what they do, and follow the basic rule of thumb that any non-trivial patch should have at least 2 or 3 paragraphs in its commit log explaining *why* the change is being made.

Quoted 3 lines

At this point, I have no idea what you are doing. I presume patches 4 and 5 are preparations for patch 6, but I don't understand what patch 6 is doing, and nowhere it is mentioned why patches 4 and 5 are needed.

Any news about these patches? From this series, we still have patches 4, 5, 6 and 8 pending in patchwork. Could you take into account the comments that have been made (mainly the strong need for more detailed commit logs), and resend updated versions on top of master?

Thanks!

Thomas

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

2014-04-19 23:58 GMT+02:00 Thomas Petazzoni
<[email protected]>:

Quoted 26 lines

Fran?ois,

Quoted 22 lines · Thomas Petazzoni
Quoted 7 lines

Francois Perrad (8):
luajit: refactor without dont-strip.patch
luajit: refactor without shared-lib.patch
luajit: handles BR2_PREFER_STATIC_LIB
luajit: add host variant
luainterpreter: add host variant
luarocks: allow to work with host-luainterpreter

For all these patches, and especially the last three ones, I want commit logs. I really do appreciate _a lot_ all the work you're doing on Lua and Perl, it's really awesome. And I really mean it. But you need to work on improving your commit logs. For now, we have no idea *why* you're doing these changes.

All the other core developers are doing efforts to write long, descriptive commit logs, and experience has shown that they are immensely useful 6 or 12 months later when we got back in the Git history to understand why things were done. Look at what they do, and follow the basic rule of thumb that any non-trivial patch should have at least 2 or 3 paragraphs in its commit log explaining *why* the change is being made.

Quoted 8 lines
Quoted 3 lines

At this point, I have no idea what you are doing. I presume patches 4 and 5 are preparations for patch 6, but I don't understand what patch 6 is doing, and nowhere it is mentioned why patches 4 and 5 are needed.

Any news about these patches? From this series, we still have patches 4, 5, 6 and 8 pending in patchwork. Could you take into account the comments that have been made (mainly the strong need for more detailed commit logs), and resend updated versions on top of master?

I'll work on them after 2014.05 release.

Fran?ois

Quoted 12 lines

Thanks!

Thomas
--
Thomas Petazzoni, CTO, Free Electrons
Embedded Linux, Kernel and Android engineering
http://free-electrons.com
_______________________________________________
buildroot mailing list
buildroot at busybox.net
http://lists.busybox.net/mailman/listinfo/buildroot

#16 lore raw

Dear Fran?ois Perrad,

Quoted 6 lines · Fran?ois Perrad
Quoted 4 lines

Any news about these patches? From this series, we still have patches 4, 5, 6 and 8 pending in patchwork. Could you take into account the comments that have been made (mainly the strong need for more detailed commit logs), and resend updated versions on top of master?

I'll work on them after 2014.05 release.

Well, there is still some time before the 2014.05 release, so you can submit them earlier if you want.

In the mean time, what can I do with your patches in patchwork? Can I mark them as "Changes requested", making the assumption you will remember to send updated versions in the future?

Thanks!

Thomas

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

2014-04-20 10:13 GMT+02:00 Thomas Petazzoni
<[email protected]>:

Quoted 9 lines

Dear Fran?ois Perrad,

Quoted 5 lines · Fran?ois Perrad
Quoted 4 lines

Any news about these patches? From this series, we still have patches 4, 5, 6 and 8 pending in patchwork. Could you take into account the comments that have been made (mainly the strong need for more detailed commit logs), and resend updated versions on top of master?

Quoted 8 lines
Quoted 1 line

I'll work on them after 2014.05 release.

Well, there is still some time before the 2014.05 release, so you can submit them earlier if you want.

In the mean time, what can I do with your patches in patchwork? Can I mark them as "Changes requested", making the assumption you will remember to send updated versions in the future?

Ok.

Fran?ois

Quoted 8 lines

Thanks!

Thomas
--
Thomas Petazzoni, CTO, Free Electrons
Embedded Linux, Kernel and Android engineering
http://free-electrons.com

#18 lore raw

v3 cover lua

In reply to Francois Perrad

Dear Francois Perrad,

Quoted 9 lines · Francois Perrad

Francois Perrad (8):
luajit: refactor without dont-strip.patch
luajit: refactor without shared-lib.patch
luajit: handles BR2_PREFER_STATIC_LIB
luajit: add host variant
luainterpreter: add host variant
luarocks: allow to work with host-luainterpreter
ljlinenoise: new package
lua-iconv: new package

From this series, there are still some patches pending in patchwork:

[lua,V3,4/8] luajit: add host variant
[lua,V3,5/8] luainterpreter: add host variant [lua,V3,6/8] luarocks: allow to work with host-luainterpreter [lua,V3,8/8] lua-iconv: new package

Could you respin them by using the host-virtual-package infrastructure and resend an updated version?

In the mean time, I'll mark those patches as "Changes Requested" in patchwork.

Thanks!

Thomas

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