I currently have ~770 hanging globus-gatekeeper processes starting from May 9.
I looked at the back-trace for an old one and a new one and saw the same thing:

(gdb) bt
#0  0x2000000000936a82 in read () from /lib/libc.so.6.1
#1  0x200000000089c8d0 in _IO_new_file_underflow () from /lib/libc.so.6.1
#2  0x200000000089f5b0 in __underflow () from /lib/libc.so.6.1
#3  0x200000000089e200 in _IO_file_xsgetn () from /lib/libc.so.6.1
#4  0x200000000089fe40 in _IO_sgetn () from /lib/libc.so.6.1
#5  0x200000000088a340 in fread () from /lib/libc.so.6.1
#6  0x2000000000056df0 in globus_gss_assist_token_get_fd
(arg=0x20000000009ebf50, bufp=0x60000fffffff9e88, sizep=0x60000fffffff9e80) at
tokens_f.c:86
#7  0x2000000000045890 in globus_gss_assist_accept_sec_context
(minor_status=0x60000fffffffa894, context_handle=0x6000000000003a50, 
    cred_handle=0x600000000001ef70, src_name_char=0x60000fffffffa8c0,
ret_flags=0x60000fffffffa89c, user_to_user_flag=0x0,
token_status=0x60000fffffffa898, 
    delegated_cred_handle=0x6000000000003a48,
gss_assist_get_token=0x2000000000a0da10,
gss_assist_get_context=0x20000000009ebf50, 
    gss_assist_send_token=0x2000000000a0da20,
gss_assist_send_context=0x6000000000022770) at accept.c:132
#8  0x4000000000009660 in doit () at globus_gatekeeper.c:1450
#9  0x4000000000008980 in main (xargc=3, xargv=0x60000fffffffbbe8) at
globus_gatekeeper.c:1124