blktap2: Fix another uninitialised value error

blktap2: Fix another uninitialised value error

gcc  -O1 -fno-omit-frame-pointer -m32 -march=i686 -g -fno-strict-aliasing -std=gnu99 -Wall -Wstrict-prototypes -Wdeclaration-after-statement   -D__XEN_TOOLS__ -MMD -MF .block-remus.o.d -D_LARGEFILE_SOURCE -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE -D_LARGEFILE64_SOURCE -fno-optimize-sibling-calls -mno-tls-direct-seg-refs -Werror -g -Wno-unused -fno-strict-aliasing -I../include -I../drivers -I/home/osstest/build.12828.build-i386/xen-unstable/tools/blktap2/drivers/../../../tools/libxc -I/home/osstest/build.12828.build-i386/xen-unstable/tools/blktap2/drivers/../../../tools/include -D_GNU_SOURCE -DUSE_NFS_LOCKS  -c -o block-remus.o block-remus.c

block-remus.c: In function 'ramdisk_flush':
block-remus.c:508: error: 'buf' may be used uninitialized in this function
make[5]: *** [block-remus.o] Error 1

This is because gcc can see that merge_requests doesn't always set
*mergedbuf but gcc isn't able to prove that it always does so if
merge_requests returns 0 and that in that case the value of
ramdisk_flush::buf isn't used.

This is too useful a warning to disable, despite the occasional false
positive of this form.  The conventional approach is to suppress the
warning by explicitly initialising the variable to 0.

This has just come to light because 25275:27d63b9f111a reenabled
optimisation for this area of code, and gcc's data flow analysis
(which is required to trigger the uninitialised variable warning) only
occurs when optimisation is turned on.

Signed-off-by: Ian Jackson <ian.jackson@eu.citrix.com>