Kevin,
Collissions are bad, but what about
IF ! STRNCMP (filename, "/proc/self/fd/", 14) THEN
fopen(filename, "r")
ELSE
fdopen(atoi(filename+14), "r")
FI
modulo better validation for the atio() arguments.
Probability of anyone using "/proc/self/fd/" as a prefix for normal
files is very small.
Small enough it will be justifiable to say "do that and you are on your own". ?
regards
ronnie sahlberg
On Thu, Jan 26, 2012 at 8:08 PM, Kevin Wolf <kw...@redhat.com> wrote:
> Am 25.01.2012 16:57, schrieb Eric Blake:
>> On 01/24/2012 11:47 PM, ronnie sahlberg wrote:
>>> Read from an arbitra filedescriptor inherited from the parent process :
>>> 9<iscsi.conf ./x86_64-softmmu/qemu-system-x86_64 -enable-kvm -display
>>> vnc=127.0.0.1:0 -drive file=iscsi://127.0.0.1/iqn.ronnie.test/1
>>> -readconfig /proc/self/fd/9
>>
>> That requires the existence of procfs, which is not portable (although
>> it does work on Linux). I'd rather see:
>>
>> -readconfig fd:9
>>
>> which matches things for -incoming; that is, if -readconfig starts with
>> '/' or '.', it is a filename; otherwise, it is a protocol:value
>> designation, where we recognize at least the fd: protocol where a value
>> is the incoming fd, but we could also recognize things like exec:
>> protocol which is an arbitrary command to use via popen.
>
> Magic prefixes like this have one big problem: What if someone has a
> config file called "fd:9"? We have the very same problem with protocols
> in the block layer and while in the general case it's a convenient
> syntax, we've come to hate it in cases where it misinterprets things.
>
> Kevin
