target/iso9660/iso9660.mk

PATCH New Roberto A. Foglietta 2 participants 7 messages Nov 21, 2008

lore

Share & tools

lore

Roberto A. Foglietta

author

2008-11-21 10:11 UTC #0 lore raw

PATCH 1/1 target/iso9660/iso9660.mk

Hi,

when the user choose to use INITRAMFS iso9660 double its size because
initrd is also copied in iso9660 but kernel has its own.

iso9660/iso9660.mk +5 -0

Index: target/iso9660/iso9660.mk
===================================================================
@@ -55,12 +55,17 @@
 endif
 
 $(ISO9660_TARGET): host-fakeroot $(LINUX_KERNEL) $(EXT2_TARGET) grub mkisofs
+	rm -rf $(ISO9660_TARGET_DIR)
 	mkdir -p $(ISO9660_TARGET_DIR)
 	mkdir -p $(ISO9660_TARGET_DIR)/boot/grub
 	cp $(GRUB_DIR)/stage2/stage2_eltorito $(ISO9660_TARGET_DIR)/boot/grub/
 	cp $(ISO9660_BOOT_MENU) $(ISO9660_TARGET_DIR)/boot/grub/menu.lst
 	cp $(LINUX_KERNEL) $(ISO9660_TARGET_DIR)/kernel
+ifneq ($(strip $(BR2_TARGET_ROOTFS_INITRAMFS)),y)
 	cp $(EXT2_TARGET) $(ISO9660_TARGET_DIR)/initrd
+else
+	touch $(ISO9660_TARGET_DIR)/initrd
+endif
 	# Use fakeroot to pretend all target binaries are owned by root
 	rm -f $(PROJECT_BUILD_DIR)/_fakeroot.$(notdir $(ISO9660_TARGET))
 	touch $(PROJECT_BUILD_DIR)/.fakeroot.00000

Signature · 8 lines

/roberto
-------------- next part --------------
A non-text attachment was scrubbed...
Name: iso9660.patch
Type: text/x-diff
Size: 823 bytes
Desc: not available
Url : http://busybox.net/lists/buildroot/attachments/20081121/1b746f4d/attachment.patch

Bernhard Reutner-Fischer

Applied

2008-11-21 10:32 UTC #1 lore raw

PATCH 1/1 target/iso9660/iso9660.mk

Quoted 20 lines · Roberto A. Foglietta

Hi,
when the user choose to use INITRAMFS iso9660 double its size because initrd is also copied in iso9660 but kernel has its own.
Index: target/iso9660/iso9660.mk
===================================================================
iso9660/iso9660.mk +0 -0
@@ -55,12 +55,17 @@
endif
$(ISO9660_TARGET): host-fakeroot $(LINUX_KERNEL) $(EXT2_TARGET) grub mkisofs
iso9660/iso9660.mk +1 -0
+	rm -rf $(ISO9660_TARGET_DIR)
	mkdir -p $(ISO9660_TARGET_DIR)
	mkdir -p $(ISO9660_TARGET_DIR)/boot/grub
	cp $(GRUB_DIR)/stage2/stage2_eltorito $(ISO9660_TARGET_DIR)/boot/grub/
	cp $(ISO9660_BOOT_MENU) $(ISO9660_TARGET_DIR)/boot/grub/menu.lst
	cp $(LINUX_KERNEL) $(ISO9660_TARGET_DIR)/kernel

Quoted 1 line

+ifneq ($(strip $(BR2_TARGET_ROOTFS_INITRAMFS)),y)

the strip is superfluous (other patches of yours also do this. Don't.)

iso9660/iso9660.mk line 67

Quoted 3 lines

	cp $(EXT2_TARGET) $(ISO9660_TARGET_DIR)/initrd

iso9660/iso9660.mk +2 -0

+else
+	touch $(ISO9660_TARGET_DIR)/initrd

Sounds like this is not needed?

I have applied something equivalent to my tree. Thanks.

Roberto A. Foglietta

author

2008-11-21 10:48 UTC #2 lore raw

PATCH 1/1 target/iso9660/iso9660.mk

2008/11/21 Bernhard Reutner-Fischer <[email protected]>:

Quoted 14 lines · Roberto A. Foglietta

Quoted 13 lines
Hi,
when the user choose to use INITRAMFS iso9660 double its size because initrd is also copied in iso9660 but kernel has its own.
Index: target/iso9660/iso9660.mk
===================================================================
iso9660/iso9660.mk (working copy) renamed from iso9660/iso9660.mk (revision 24103) +0 -0
@@ -55,12 +55,17 @@
endif

Quoted 11 lines

iso9660/iso9660.mk (working copy)
Quoted 8 lines
$(ISO9660_TARGET): host-fakeroot $(LINUX_KERNEL) $(EXT2_TARGET) grub mkisofs
iso9660/iso9660.mk (working copy) renamed from iso9660/iso9660.mk (revision 24103) +2 -0
+      rm -rf $(ISO9660_TARGET_DIR)
      mkdir -p $(ISO9660_TARGET_DIR)
      mkdir -p $(ISO9660_TARGET_DIR)/boot/grub
      cp $(GRUB_DIR)/stage2/stage2_eltorito $(ISO9660_TARGET_DIR)/boot/grub/
      cp $(ISO9660_BOOT_MENU) $(ISO9660_TARGET_DIR)/boot/grub/menu.lst
      cp $(LINUX_KERNEL) $(ISO9660_TARGET_DIR)/kernel
+ifneq ($(strip $(BR2_TARGET_ROOTFS_INITRAMFS)),y)
the strip is superfluous (other patches of yours also do this. Don't.)

Ok, I just imitate the others code.

Quoted 6 lines

iso9660/iso9660.mk (working copy)
Quoted 3 lines
      cp $(EXT2_TARGET) $(ISO9660_TARGET_DIR)/initrd
iso9660/iso9660.mk (working copy) renamed from iso9660/iso9660.mk (revision 24103) +2 -0
+else
+      touch $(ISO9660_TARGET_DIR)/initrd
Sounds like this is not needed?

I strongly suggest to add the touch initrd because:

a) initramf takes a lot of time at each compilation so beta-testing
stage usually goes with initrd

b) when you deliver the product you should turn on initramfs but then
menu.lst still reports grub should load initrd but if it does not exist grub fails

touching the initrd means: if you forget to modify menu.lst you will
not pay the fee to see your system unable to boot.

The question: is there any condition in which touching initrd really
hurts somebody? I think the answer is no.

Under this point of view touching initrd save time to skilled users
and make things work for those are not such skilled users

Cheers,

Signature · 1 line

/roberto

Bernhard Reutner-Fischer

2008-11-21 11:11 UTC #3 lore raw

PATCH 1/1 target/iso9660/iso9660.mk

Quoted 27 lines · Roberto A. Foglietta

2008/11/21 Bernhard Reutner-Fischer <[email protected]>:
Quoted 14 lines · Roberto A. Foglietta
Quoted 13 lines
Hi,
when the user choose to use INITRAMFS iso9660 double its size because initrd is also copied in iso9660 but kernel has its own.
Index: target/iso9660/iso9660.mk
===================================================================
iso9660/iso9660.mk (working copy) renamed from iso9660/iso9660.mk (revision 24103) +0 -0
@@ -55,12 +55,17 @@
endif
Quoted 11 lines
iso9660/iso9660.mk (working copy)
Quoted 8 lines
$(ISO9660_TARGET): host-fakeroot $(LINUX_KERNEL) $(EXT2_TARGET) grub mkisofs
iso9660/iso9660.mk (working copy) renamed from iso9660/iso9660.mk (revision 24103) +2 -0
+      rm -rf $(ISO9660_TARGET_DIR)
      mkdir -p $(ISO9660_TARGET_DIR)
      mkdir -p $(ISO9660_TARGET_DIR)/boot/grub
      cp $(GRUB_DIR)/stage2/stage2_eltorito $(ISO9660_TARGET_DIR)/boot/grub/
      cp $(ISO9660_BOOT_MENU) $(ISO9660_TARGET_DIR)/boot/grub/menu.lst
      cp $(LINUX_KERNEL) $(ISO9660_TARGET_DIR)/kernel
+ifneq ($(strip $(BR2_TARGET_ROOTFS_INITRAMFS)),y)
the strip is superfluous (other patches of yours also do this. Don't.)

Quoted 1 line

Ok, I just imitate the others code.

Yea, but please do not immitate bloat..

Quoted 17 lines

Quoted 7 lines
iso9660/iso9660.mk (working copy)
Quoted 4 lines
      cp $(EXT2_TARGET) $(ISO9660_TARGET_DIR)/initrd
iso9660/iso9660.mk (working copy) renamed from iso9660/iso9660.mk (revision 24103) +2 -0
+else
+      touch $(ISO9660_TARGET_DIR)/initrd
Sounds like this is not needed?
I strongly suggest to add the touch initrd because:
a) initramf takes a lot of time at each compilation so beta-testing stage usually goes with initrd
b) when you deliver the product you should turn on initramfs but then menu.lst still reports grub should load initrd but if it does not exist grub fails

Quoted 5 lines

touching the initrd means: if you forget to modify menu.lst you will not pay the fee to see your system unable to boot.
The question: is there any condition in which touching initrd really hurts somebody? I think the answer is no.

Yes, it hurts since it's an unwarranted inode and adds to the size of the image needlessly. It is simply not clean and not the proper thing to do.

Quoted 3 lines

Under this point of view touching initrd save time to skilled users and make things work for those are not such skilled users

I did it a bit differently, fyi.

Roberto A. Foglietta

author

2008-11-21 11:22 UTC #4 lore raw

PATCH 1/1 target/iso9660/iso9660.mk

2008/11/21 Bernhard Reutner-Fischer <[email protected]>:

Quoted 3 lines · Roberto A. Foglietta

Quoted 2 lines
2008/11/21 Bernhard Reutner-Fischer <[email protected]>:
Quoted 1 line · Roberto A. Foglietta

[cut]

Quoted 6 lines

Quoted 5 lines
touching the initrd means: if you forget to modify menu.lst you will not pay the fee to see your system unable to boot.
The question: is there any condition in which touching initrd really hurts somebody? I think the answer is no.

Quoted 9 lines

Yes, it hurts since it's an unwarranted inode and adds to the size of the image needlessly. It is simply not clean and not the proper thing to do.
Quoted 3 lines
Under this point of view touching initrd save time to skilled users and make things work for those are not such skilled users
I did it a bit differently, fyi.

I happy to know there is a better way to do that. Could you post here to for the sake of completeness, please?

Ciao,
--
/roberto

Bernhard Reutner-Fischer

2008-11-21 11:50 UTC #5 lore raw

PATCH 1/1 target/iso9660/iso9660.mk

Quoted 5 lines · Roberto A. Foglietta

Quoted 2 lines
I did it a bit differently, fyi.
I happy to know there is a better way to do that. Could you post here to for the sake of completeness, please?

http://repo.or.cz/w/buildroot.git?a=commit;h=c5e0d6d2ebdfb4cfcde877b73fa88538644b4ece
http://repo.or.cz/w/buildroot.git?a=commit;h=e25696afb0d450e5c7bc7c4be2e2a4275c5722f3
http://repo.or.cz/w/buildroot.git?a=commit;h=7be47a7657c7aa929cda3f57a6a1102bd4a3d3c1

Roberto A. Foglietta

author

2008-11-21 12:04 UTC #6 lore raw

PATCH 1/1 target/iso9660/iso9660.mk

2008/11/21 Bernhard Reutner-Fischer <[email protected]>:

Quoted 5 lines · Roberto A. Foglietta

Quoted 4 lines
Quoted 3 lines
I did it a bit differently, fyi.

Quoted 7 lines

Quoted 2 lines
I happy to know there is a better way to do that. Could you post here to for the sake of completeness, please?
http://repo.or.cz/w/buildroot.git?a=commit;h=c5e0d6d2ebdfb4cfcde877b73fa88538644b4ece
http://repo.or.cz/w/buildroot.git?a=commit;h=e25696afb0d450e5c7bc7c4be2e2a4275c5722f3
http://repo.or.cz/w/buildroot.git?a=commit;h=7be47a7657c7aa929cda3f57a6a1102bd4a3d3c1

Yes, you are right: only copied version of menu.lst should be modified. :-)

Ciao,

Signature · 1 line

/roberto