"cbb0: bad Vcc request"?

[der Mouse <mouse%Rodents-Montreal.ORG@localhost>]

I have an i386 laptop which has a PCMCIA socket (or something that
looks like it to me; I know there are at least two variants, subtly
different, and I am not competent to tell the difference between them
without help).  However, it doesn't work.  No matter what I plug into
it, the kernel remains sublimely oblivious.

I suspect the reason is related to some boot-time messages.  The
machine is running 4.0.1; here's dmesg | grep cbb from a boot done with
nothing physically present in the socket:

cbb0 at pci0 dev 10 function 0: Texas Instruments PCI4410 PCI-CardBus Bridge w/ 
IEEE 1394 Host Controller (rev. 0x02)
cbb0: bad Vcc request. sock_ctrl 0x0, sock_status 0xf000e2c3
cbb0: disabling socket
cbb0: interrupting at irq 9
cbb0: secondary bus number uninitialized; try PCI_BUS_FIXUP
cardslot0 at cbb0 slot 0 flags 0

That "bad Vcc request" line looks suspicious.

Looking at the code, this appears to be coming from the hardware, but
it's not at all clear to me what it means.  "Bad Vcc Request" could
mean anything from "a card is requesting something impossible" to "the
host CPU just asked me to do something I can't" to "a selftest failed".

To save people looking it up, f000e2c3 is

CB_SOCKET_STAT_YVSTAT   set
CB_SOCKET_STAT_XVSTAT   set
CB_SOCKET_STAT_3VSTAT   set
CB_SOCKET_STAT_5VSTAT   set
CB_SOCKET_STAT_YVCARD   set
CB_SOCKET_STAT_XVCARD   clear
CB_SOCKET_STAT_3VCARD   clear
CB_SOCKET_STAT_5VCARD   clear
CB_SOCKET_STAT_BADVCC   set (of course)
CB_SOCKET_STAT_DATALOST clear
CB_SOCKET_STAT_NOTCARD  set
CB_SOCKET_STAT_IREQ     set
CB_SOCKET_STAT_CB       clear
CB_SOCKET_STAT_16BIT    clear
CB_SOCKET_STAT_PWRCYCLE clear
CB_SOCKET_STAT_CD2      clear
CB_SOCKET_STAT_CD1      set
CB_SOCKET_STAT_CARDSTS  set
0x000c0000 (for which there's no #define in pccbbreg.h) set

The presence of NOTCARD ("Inserted card is unrecognisable") and CD1
("card detect 1"), when there actually is no card inserted, make me
suspect that the hardware is broken.  Is this a reasonable inference?
If not, any ideas what could be going on?

/~\ The ASCII                             Mouse
\ / Ribbon Campaign
 X  Against HTML                mouse%rodents-montreal.org@localhost
/ \ Email!           7D C8 61 52 5D E7 2D 39  4E F1 31 3E E8 B3 27 4B