make3-82/linux-2.6.28 Dependencies recipe running (THIS IS A BUG)

[John Breitenbach]

I've encountered a problem with building the linux kernel v2.6.28 with 
make-3.82.   This
has been building fine with version 3.81 for some time.  And while 
attempting to diagnose
the problem, I added "-p" to the make command line encountered the "THIS 
IS A BUG" message.
This is running on a 32-bit RedHat EL5 system.  Number of parallel jobs 
has no effect.

The original error message from building the kernel is as follows:

make[2]: *** No rule to make target `firmware/edgeport/boot.fw', needed 
by `__fw_modbuild'.  Stop.
make[2]: *** Waiting for unfinished jobs....

Here's part of firmware/Makefile from linux v2.6.28.  It contains two 
rules to build a .fw file, and
there IS a .SECONDEXPANSION: earlier in this Makefile:

# .HEX is also Intel HEX, but where the offset and length in each record
# is actually meaningful, because the firmware has to be loaded in a certain
# order rather than as a single binary blob. Thus, we convert them into our
# more compact binary representation of ihex records (<linux/ihex.h>)
$(obj)/%.fw: $(obj)/%.HEX $(ihex2fw_dep) | $(objtree)/$(obj)/$$(dir %)
   $(call cmd,ihex2fw)

# .H16 is our own modified form of Intel HEX, with 16-bit length for 
records.
$(obj)/%.fw: $(obj)/%.H16 $(ihex2fw_dep) | $(objtree)/$(obj)/$$(dir %)
   $(call cmd,h16tofw)

$(firmware-dirs):
   $(call cmd,mkdir)

There is an order-only dependency to create the necessary subdirectory.  
And while this Makefile contains
a rule to make subdirectories, the "firmware-dirs" variable contains TWO 
trailing "/" in each of its elements.

Here's a couple more lines from firmware/Makefile which sets the 
firmware-dirs variable.

# Directories which we _might_ need to create, so we have a rule for them.
firmware-dirs := $(sort $(patsubst %,$(objtree)/$(obj)/%/,$(dir 
$(fw-external-y) $(fw-shipped-all))))

If I remove the extra slash right before the comma, the build succeeds, 
and the error
message does not appear.  I also have found that if the firmware/edgport 
directory exists,
the entire build succeeds.

I know I'm not giving you guys much to work with - perhaps you can spot 
the error in ensuring
the presence of the output directory.    Or perhaps I can get some 
advice finding the cause of the
"THIS IS A BUG" message.

John