linux-headers upgrade to 3.0.39

[email protected]
Newsgroups gmane.comp.security.ipcop.devel
Message-ID <226219656.91046677.1343987799563.JavaMail.root@zimbra33-e6.priv.proxad.net>
I tried Olaf suggestion to upgrade the kernel headers and result look good.

A bare upgrade break :
- udev
I backported to 166 the videodev change.
167 behave differently with --disable-extras and no more build any extra (like input_id, scsi_id, usb_id)
that could be reverted on 168 but I haven't yet explored what was broken there.

- ppp
A simple 'cd $(DIR_APP) && rm -f include/linux/if_pppol2tp.h' unbreak that build like veryone do.

- busybox
I borrowed from master branch ubi __packed change to 1.19.4, maybe 1.20.2 work, untested yet.
This suggest too more ubi could be disabled in our busybox config as .config diff from 1.17 to .19 show
 # CONFIG_UBIATTACH is not set
 # CONFIG_UBIDETACH is not set
+CONFIG_UBIMKVOL=y
+CONFIG_UBIRMVOL=y
+CONFIG_UBIRSVOL=y
+CONFIG_UBIUPDATEVOL=y
 # CONFIG_ADJTIMEX is not set
 # CONFIG_BBCONFIG is not set


The change in the kernel upgrade could be seen in log as :
diff -ur log_i486-2012-08-02-1/02_base/util-linux-2.21.2 log_i486-2012-08-02-2/02_base/util-linux-2.21.2
--- log_i486-2012-08-02-1/02_base/util-linux-2.21.2	2012-08-02 14:36:29.000000000 +0200
+++ log_i486-2012-08-02-2/02_base/util-linux-2.21.2	2012-08-03 01:52:34.000000000 +0200
@@ -236,7 +236,7 @@
 checking for langinfo.h... yes
 checking for struct termios.c_line... yes
 checking for struct stat.st_mtim.tv_nsec... yes
-checking whether UNAME26 is declared... no
+checking whether UNAME26 is declared... yes
 checking whether ADDR_NO_RANDOMIZE is declared... yes
 checking whether FDPIC_FUNCPTRS is declared... yes
 checking whether MMAP_PAGE_ZERO is declared... yes
@@ -326,9 +326,8 @@
 checking for syscall unshare... SYS_unshare
 checking for unshare... yes
 checking for crc32 in -lz... yes
-checking for syscall prlimit64... no
-configure: WARNING: Unable to detect syscall prlimit64.
-configure: WARNING: prlimit64 syscall not found; do not build prlimit
+checking for syscall prlimit64... SYS_prlimit64
+checking for prlimit... no
 checking for syscall pivot_root... SYS_pivot_root
 configure: WARNING: schedutils disabled; do not build ionice
 configure: WARNING: schedutils disabled; do not build taskset
...
+libtool: install: /tools_i486/bin/install -c prlimit /usr/bin/prlimit



--- log_i486-2012-08-02-1/03_ipcop/libpcap-1.3.0	2012-08-02 15:10:41.000000000 +0200
+++ log_i486-2012-08-02-2/03_ipcop/libpcap-1.3.0	2012-08-03 02:28:42.000000000 +0200
@@ -128,7 +128,7 @@
 checking for linux/usbdevice_fs.h... yes
 checking if usbdevfs_ctrltransfer struct has bRequestType member... yes
 checking whether the platform could support netfilter sniffing... yes
-checking whether we can compile the netfilter support... no
+checking whether we can compile the netfilter support... yes
 checking bluetooth/bluetooth.h usability... no
 checking bluetooth/bluetooth.h presence... no
 checking for bluetooth/bluetooth.h... no
@@ -137,9 +137,9 @@
 checking libusb-1.0/libusb.h presence... yes
 checking for libusb-1.0/libusb.h... yes
 configure: canusb sniffing is supported
-checking linux/net_tstamp.h usability... no
-checking linux/net_tstamp.h presence... no
-checking for linux/net_tstamp.h... no
+checking linux/net_tstamp.h usability... yes
+checking linux/net_tstamp.h presence... yes
+checking for linux/net_tstamp.h... yes
...
+gcc -O2 -fpic -Os -march=i486 -mtune=pentium -pipe -fomit-frame-pointer -I.  -DHAVE_CONFIG_H  -D_U_="__attribute__((unused))" -Os -march=i486 -mtune=pentium -pipe -fomit-frame-pointer -c ./pcap-netfilter-linux.c

This could be a regression or may that be intented?
--- log_i486-2012-08-02-1/03_ipcop/setserial-2.17	2012-08-02 16:01:03.000000000 +0200
+++ log_i486-2012-08-02-2/03_ipcop/setserial-2.17	2012-08-03 03:21:00.000000000 +0200
@@ -26,14 +26,14 @@
 checking for strip... strip
 checking how to run the C preprocessor... gcc -E
 checking for asm/ioctls.h... yes
-checking for linux/hayesesp.h... yes
+checking for linux/hayesesp.h... no
 updating cache ./config.cache
 creating ./config.status
 creating setserial.8
 creating Makefile
 cd /usr/src/setserial-2.17 && make
 make[1]: Entering directory `/usr/src/setserial-2.17'
-gcc -Os -march=i486 -mtune=pentium -pipe -fomit-frame-pointer -Wall -DHAVE_ASM_IOCTLS_H=1 -DHAVE_LINUX_HAYESESP_H=1  -I. setserial.c -o setserial
+gcc -Os -march=i486 -mtune=pentium -pipe -fomit-frame-pointer -Wall -DHAVE_ASM_IOCTLS_H=1  -I. setserial.c -o setserial


There is not a lot of more different files with the kernel upgrade
diff -u doc/IPCop-2.1.0-diff-list-2012-08-02*
--- doc/IPCop-2.1.0-diff-list-2012-08-02-1.i486.txt     2012-08-02 16:18:59.000000000 +0200
+++ doc/IPCop-2.1.0-diff-list-2012-08-02-2.i486.txt     2012-08-03 03:39:51.000000000 +0200
@@ -21,9 +21,14 @@
 DIFF    /lib/libpthread-2.11.3.so                                    glibc
 DIFF    /lib/libresolv-2.11.3.so                                     glibc
 DIFF    /lib/librt-2.11.3.so                                         glibc
+DIFF    /lib/libudev.so.0.10.0                                       udev
 DIFF    /lib/libutil-2.11.3.so                                       glibc
+DIFF    /lib/udev/ata_id                                             udev
 DIFF    /sbin/extlinux                                               syslinux
 DIFF    /sbin/ldconfig                                               glibc
+DIFF    /sbin/udevadm                                                udev
+DIFF    /sbin/udevd                                                  udev
+DIFF    /usr/bin/fgconsole                                           kbd
 DIFF    /usr/bin/ntpd                                                ntp
 DIFF    /usr/bin/ntpdate                                             ntp
 DIFF    /usr/bin/ntpq                                                ntp
@@ -33,6 +38,7 @@
 DIFF    /usr/lib/libdb-4.8.so                                        db
 DIFF    /usr/lib/libgcc_s.so.1                                       gcc
 DIFF    /usr/lib/libhistory.so.6.2                                   readline
+DIFF    /usr/lib/libnfnetlink.so.0.2.0                               libnfnetlink
 DIFF    /usr/lib/libreadline.so.6.2                                  readline
 DIFF    /usr/lib/perl5/5.10.1/i486-linux/auto/Compress/Raw/Zlib/Zlib.so perl
 DIFF    /usr/lib/perl5/5.10.1/i486-linux/auto/Socket/Socket.so       perl
@@ -45,6 +51,7 @@
 DIFF    /usr/lib/pppd/2.4.5/rp-pppoe.so                              ppp
 DIFF    /usr/sbin/chat                                               ppp
 DIFF    /usr/sbin/iftop                                              iftop
+DIFF    /usr/sbin/iwevent                                            wireless_tools
 DIFF    /usr/sbin/parted                                             parted
 DIFF    /usr/sbin/partprobe                                          parted
 DIFF    /usr/sbin/pppd                                               ppp
@@ -574,6 +581,7 @@
 UPDATE  /usr/sbin/useradd
 UPDATE  /usr/sbin/userdel
 UPDATE  /usr/sbin/usermod
+UPDATE  /usr/sbin/wanpipemon
 UPDATE  /usr/share/locale/af_ZA/LC_MESSAGES/ipcop.mo
 UPDATE  /usr/share/locale/bg_BG/LC_MESSAGES/ipcop.mo
 UPDATE  /usr/share/locale/ca_ES/LC_MESSAGES/ipcop.mo

So I would conclude upgrading the kernel headers to 3.0.39 is a good thing to do.
(linux-3.0 headers are not fixed for the parted issue on include/linux/ext2_fs.h)

Any other opinion?

Concerning udev and busybox, we may upgrade but that should be another subject.

Gilles



------------------------------------------------------------------------------
Live Security Virtual Conference
Exclusive live event will cover all the ways today's security and 
threat landscape has changed and how IT managers can respond. Discussions 
will include endpoint security, mobile security and the latest in malware 
threats. http://www.accelacomm.com/jaw/sfrnl04242012/114/50122263/
lmpx.com only provides a reader for public news (NNTP) servers. It is not affiliated with the servers or forums shown here and is not responsible for the content of articles, which is written by their respective authors.