Re: [Xen-devel] [PATCH v5 03/21] acpi: Prevent GPL-only code from seeping into non-GPL binaries

[Ian Jackson <ian.jackson@xxxxxxxxxxxxx>]

Boris Ostrovsky writes ("Re: [PATCH v5 03/21] acpi: Prevent GPL-only code from 
seeping into non-GPL binaries"):
> >> +printf "\n    /* Beginning of GPL-only code */\n\n"
> >> +
> >> +printf "    /* _S3 and _S4 are in separate SSDTs */\n"
> >> +printf "    Name (\_S5, Package (0x04) {\n"
> >> +printf "        0x00,  /* PM1a_CNT.SLP_TYP */\n"
> >> +printf "        0x00,  /* PM1b_CNT.SLP_TYP */\n"
> >> +printf "        0x00,  /* reserved */\n"
> >> +printf "        0x00   /* reserved */\n"
> >> +printf "    })\n"
> >> +
> >> +printf "    Name(PICD, 0)\n"
> >> +printf "    Method(_PIC, 1) {\n"
> >> +printf "        Store(Arg0, PICD)\n"
> >> +printf "    }\n"
> > Wouldn't this be better readable with "echo", avoiding all the \n
> > instances? Actually this seems to even apply to most if not
> > everything further down, as so far I didn't spot a case where
> > you actually pass anything other than just a format string.
> 
> Format string is the only reason. There are a couple of instances where
> output is formatted and I felt that having both echo and printf would
> feel inconsistent.

I don't think there is anything wrong with inconsistency.

But really, why not make this (or most of it) a here document (ie with
<<) ?  That would remove a lot of the quoting clutter (and would also
avoid any bugs where shell metacharacters or printf metacharacters are
unintentionally written unquoted).

Ian.

_______________________________________________
Xen-devel mailing list
Xen-devel@xxxxxxxxxxxxx
https://lists.xen.org/xen-devel