Module name --> Place in Kernel

[Rizado]

There's something I've always wanted to know. When I compile a new kernel I look what modules are loaded by lsmod to see what I have. But often the module name is not used in the kernel description.

Module Size Used by
nf_nat_ftp 3296 0
nf_conntrack_ftp 8896 1 nf_nat_ftp
nf_nat_irc 2688 0
nf_conntrack_irc 6648 1 nf_nat_irc
ipt_LOG 6080 12
xt_limit 2496 14
ipt_MASQUERADE 3328 1
iptable_nat 7268 1
nf_nat 19212 4 nf_nat_ftp,nf_nat_irc,ipt_MASQUERADE,iptable_nat
xt_length 1888 3
xt_state 2400 25
xt_MARK 2240 19
iptable_mangle 2688 1
iptable_filter 2848 1
nf_conntrack_ipv4 17736 27 iptable_nat
nf_conntrack 64204 9 nf_nat_ftp,nf_conntrack_ftp,nf_nat_irc,nf_conntrac k_irc,ipt_MASQUERADE,iptable_nat,nf_nat,xt_state,n f_conntrack_ipv4
ip_tables 12904 3 iptable_nat,iptable_mangle,iptable_filter
act_police 5924 1
cls_u32 7748 4
sch_ingress 3648 1
cls_fw 5120 4
sch_sfq 6016 4
sch_htb 17632 1
loop 17060 0
pcspkr 2752 0
evdev 10176 0
fuse 44500 1

So how do I know what options to keep/enable in the kernel config? Some things like fuse and loop is pretty obvious but all nat stuff is, well...

[kidders]

Hi there,

Most of the time, the kernel configurator should tell you the names of modules you are building. For example, choosing one of the modules you listed at random, if you select (in a recent kernel) Networking -> Networking support -> Networking options -> QoS and/or fair queueing -> Ingress Qdisc, you should see "To compile this code as a module, choose M here: the module will be called sch_ingress."

Also, module names are often very closely related to the names of the kernel config options that govern them. For instance ...# find /usr/src/linux/ -name Makefile -exec grep sch_ingress {} \;
obj-$(CONFIG_NET_SCH_INGRESS) += sch_ingress.o... essentially tells you that the kernel option you need to enable to keep the "sch_ingress" module is called NET_SCH_INGRESS, but that would have been easy to guess without searching the source tree for help.

Since you mentioned NAT, I should perhaps have chosen a slightly better example ... say maybe packet mangling (which controls your firewall's ability to modify packets passing through it). It's kernel config option is called IP_NF_MANGLE. To check for packet mangling support, you could ...# See if the kernel you're currently running has support for it...
$ grep IP_NF_MANGLE /boot/config-`uname -r`
CONFIG_IP_NF_MANGLE=m

# Check that the kernel you've just configured does...
$ grep IP_NF_MANGLE /usr/src/linux/.config
CONFIG_IP_NF_MANGLE=mIf not, find "Packet mangling" in your configurator, and you should see something along the lines of "Packet mangling (IP_NF_MANGLE) This option adds a `mangle' table to iptables...", so you know you're enabling the right thing.

I hope that helps.