Linux問題情報分享（3）：CentOS7上最新kernel-debuginfo包與當前核心版本不匹配

CentOS 7上最新的kernel-debuginfo包，是kernel-debuginfo-4.x.x-x.el7，而當前核心是kernel-3.10.0。因此，如果安裝或者升級到了最新的kernel-debuginfo包，會導致類似SystemTap這樣需要核心標頭檔案和除錯符號的工具執行出錯。這是CentOS 7的bug。

以SystemTap為例，其報錯大致如下

[root@pusf ~]# rpm -qa | grep kernel-debuginfo
kernel-debuginfo-4.9.31-203.el7.centos.x86_64
kernel-debuginfo-common-x86_64-4.9.31-203.el7.centos.x86_64[root@pusf ~]
[root@pusf ~]# stap -v -e `probe vfs.read {printf("read performed
"); exit()}`
Pass 1: parsed user script and 119 library scripts using 117224virt/33688res/3120shr/30824data kb, in 240usr/10sys/292real ms.
semantic error: while resolving probe point: identifier `kernel` at /usr/share/systemtap/tapset/linux/vfs.stp:882:18
        source: probe vfs.read = kernel.function("vfs_read")
                      ^
semantic error: missing x86_64 kernel/module debuginfo [man warning::debuginfo] under `/lib/modules/3.10.0-514.26.2.el7.x86_64/build`
semantic error: while resolving probe point: identifier `vfs` at <input>:1:7
        source: probe vfs.read {printf("read performed
"); exit()}
                      ^
semantic error: no match

Pass 2: analyzed script: 0 probes, 0 functions, 0 embeds, 0 globals using 120568virt/37092res/5108shr/32068data kb, in 100usr/140sys/443real ms.
Missing separate debuginfos, use: debuginfo-install kernel-3.10.0-514.26.2.el7.x86_64
Pass 2: analysis failed.  [man error::pass2]
[root@pusf ~]#

因此，出現這種情況時，需要解除安裝kernel-debuginfo-4.x.x-x.el7和kernel-debuginfo-common-4.x.x-x.el7的包，重新按照當前核心版本安裝kernel-debuginfo即可

rpm -qa | grep -E `^kernel-` | grep -v 3.10.0 | xargs yum -y remove
debuginfo-install -y kernel-$(uname -r)

再測試下SystemTap的指令碼，會發現問題已經解決了

[root@pusf ~]# rpm -qa | grep kernel-debuginfo                    
kernel-debuginfo-common-x86_64-3.10.0-514.26.2.el7.x86_64
kernel-debuginfo-3.10.0-514.26.2.el7.x86_64
[root@pusf ~]# stap -v -e `probe vfs.read {printf("read performed
"); exit()}`

Pass 1: parsed user script and 119 library scripts using 117224virt/33684res/3120shr/30824data kb, in 230usr/10sys/293real ms.

Pass 2: analyzed script: 1 probe, 1 function, 4 embeds, 0 globals using 248216virt/165932res/4336shr/161816data kb, in 1260usr/370sys/1809real ms.

Pass 3: translated to C into "/tmp/stapI5iwL4/stap_1aa47f863c3f13e51da3e80cc92942be_1682_src.c" using 248216virt/166236res/4640shr/161816data kb, in 20usr/40sys/57real ms. 

Pass 4: compiled C into "stap_1aa47f863c3f13e51da3e80cc92942be_1682.ko" in 5550usr/1240sys/7187real ms.

Pass 5: starting run.
read performed
Pass 5: run completed in 0usr/60sys/378real ms.
[root@pusf ~]#

在CentOS官方修正bug前，可以在/etc/yum.conf中加入如下配置，先排除問題包

exclude=kernel-debuginfo*

這樣，升級時不會再次安裝了問題包。當然，核心升級時，需要額外調整下配置。

順便說下，在Ubuntu上使用SystemTap，需要額外配置和步驟，請參考SystemTap on Ubuntu。

參考

1. SystemTap
2. SystemTap on CentOS
3. SystemTap on Ubuntu
4. Where to find the kernel-debuginfo package