一个 perl 处理文本的问题

[terry]

要处理一个 XenServer dom0 崩溃的问题，感觉是硬件问题（应该是处理器）。需要查 MCE 错误代码，但问题是 XenServer dom0 (CentOS 5.x) kern.log 的格式是非标准的，用 mcelog 程序没法直接读，需要处理一下，查到一个 perl 1-liner 但本人对 perl 一无所知，调试了一下无果，错误看不懂。

cat kern.log | perl -ne '/CPU(\d+), BANK(\d+), addr (\S+), state (\S+)/; print "CPU $1 BANK $2 TSC 00000000000000\nMISC 0000000000000000 ADDR $3 \nSTATUS $4\n" > mce_fmtd.log'
Can't take log of 0 at -e line 1, <> line 1.

请问一下，有没有办法用 bash 脚本，利用 awk, sed, GNU 命令行工具处理得到同样的结果呢？

kern.log 的格式如下

Apr 16 16:42:10 server109 kernel: MCE_DOM0_LOG: enter dom0 mce vIRQ handler
Apr 16 16:42:10 server109 kernel: MCE_DOM0_LOG: No more urgent data
Apr 16 16:42:10 server109 kernel: [CPU8, BANK0, addr db484ffc0, state cc0000c00001009f]
Apr 16 16:42:10 server109 kernel: MCE_DOM0_LOG: No more nonurgent data
Apr 16 16:42:10 server109 kernel: MCE_DOM0_LOG: enter dom0 mce vIRQ handler
Apr 16 16:42:10 server109 kernel: MCE_DOM0_LOG: No more urgent data
Apr 16 16:42:10 server109 kernel: [CPU8, BANK0, addr db484ffc0, state cc0005000001009f]
Apr 16 16:42:10 server109 kernel: MCE_DOM0_LOG: No more nonurgent data
Apr 16 16:42:26 server109 kernel: MCE_DOM0_LOG: enter dom0 mce vIRQ handler
Apr 16 16:42:26 server109 kernel: MCE_DOM0_LOG: No more urgent data
Apr 16 16:42:26 server109 kernel: [CPU8, BANK0, addr db484fcc0, state cc0004c00001009f]
Apr 16 16:42:26 server109 kernel: MCE_DOM0_LOG: No more nonurgent data
Apr 16 16:42:57 server109 kernel: MCE_DOM0_LOG: enter dom0 mce vIRQ handler
Apr 16 16:42:57 server109 kernel: MCE_DOM0_LOG: No more urgent data
Apr 16 16:42:57 server109 kernel: [CPU8, BANK0, addr db484ef40, state cc0002800001009f]
Apr 16 16:42:57 server109 kernel: MCE_DOM0_LOG: No more nonurgent data

[11138]

cat kern.log | perl -ne '/CPU(\d+), BANK(\d+), addr (\S+), state (\S+)/; print "CPU $1 BANK $2 TSC 00000000000000\nMISC 0000000000000000 ADDR $3 \nSTATUS $4\n"' > mce_fmtd.log

[terry]

其实我只要知道

cat kern.log | perl -ne '/CPU(\d+), BANK(\d+), addr (\S+), state (\S+)/; print "CPU $1 BANK $2 TSC 00000000000000\nMISC 0000000000000000 ADDR $3 \nSTATUS $4\n"

是要把下面的文本弄成什么样的目标格式就行了

CPU8, BANK0, addr db484ffc0, state cc0000c00001009f

转成

CPU 8, BANK 0, TSC 00000000000000
MISC 0000000000000000 ADDR $3
STATUS $4

这样么？

谢谢先

[terry]

@11138 谢了，仔细看后就发现问题了。

下班前果然没心思看，竟然没发现单引号把重定向符号给包起来了...

[terry]

给感兴趣的人看一下结果

decode 之后

sudo mcelog --cpu=westmere --ascii --no-dmi --ignorenodev --file mce_fmtd.log > mce_decoded.log

CPU 8 BANK 0 TSC 00000000000000
MISC 0000000000000000 ADDR db484ef40
Hardware event. This is not a software error.
CPU 8 BANK 0
MISC 0 ADDR db484ef40
MCG status:
MCi status:
Error overflow
Corrected error
MCi_MISC register valid
MCi_ADDR register valid
MCA: MEMORY CONTROLLER RD_CHANNELunspecified_ERR
Transaction: Memory read error
Memory read ECC error
Memory corrected error count (CORE_ERR_CNT): 10
Memory transaction Tracker ID (RTId): 0
Memory DIMM ID of error: 0
Memory channel ID of error: 0
Memory ECC syndrome: 0
STATUS cc0002800001009f MCGSTATUS 0
STATUS cc0002800001009f]
CPU 8 BANK 0 TSC 00000000000000
MISC 0000000000000000 ADDR db484ef40
Hardware event. This is not a software error.

知道的一看就明白是什么问题了...

[ngn999]

cat kern.log | perl -ne 'print "CPU $1 BANK $2 TSC 00000000000000\nMISC 0000000000000000 ADDR $3 \nSTATUS $4\n" if /CPU(\d+), BANK(\d+), addr (\S+), state (\S+)/' > mce_fmtd.log

[ngn999]

perl在命令行下当 sed + awk用

[leoYu]

用sed也可以解决的，lz可以邮件我