Joshua Oreman [Mon, 10 Aug 2009 05:12:21 +0000 (22:12 -0700)]
[zbin] Change fixup semantics to support ROMs over 128k uncompressed
The option ROM header contains a one-byte field indicating the number
of 512-byte sectors in the ROM image. Currently it is linked to
contain the number of uncompressed sectors, with an instruction to the
compressor to correct it. This causes link failure when the
uncompressed size of the ROM image is over 128k.
Fix by replacing the SUBx compressor fixup with an ADDx fixup that
adds the total compressed output length, scaled as requested, to an
addend stored in the field where the final length value will be
placed. This is similar to the behavior of ELF relocations, and
ensures that an overflow error will not be generated unless the
compressed size is still too large for the field.
This also allows us to do away with the _filesz_pgh and _filesz_sect
calculations exported by the linker script.
Output tested bitwise identical to the old SUBx mechanism on hd, dsk,
lkrn, and rom prefixes, on both 32-bit and 64-bit processors.
Modified-by: Michael Brown <mcb30@etherboot.org> Signed-off-by: Michael Brown <mcb30@etherboot.org>
Michael Brown [Mon, 10 Aug 2009 20:22:31 +0000 (21:22 +0100)]
[infiniband] Add support for the SRP Boot Firmware Table
The SRP Boot Firmware Table serves a similar role to the iSCSI and AoE
Boot Firmware Tables; it provides information required by the loaded
OS in order to establish a connection back to the SRP boot device.
Michael Brown [Sun, 9 Aug 2009 20:00:48 +0000 (21:00 +0100)]
[infiniband] Disambiguate CM connection rejection reasons
There is diagnostic value in being able to disambiguate between the
various reasons why an IB CM has rejected a connection attempt. In
particular, reason 8 "invalid service ID" can be used to identify an
incorrect SRP service_id root-path component, and reason 28 "consumer
reject" corresponds to a genuine SRP login rejection IU, which can be
passed up to the SRP layer.
For rejection reasons other than "consumer reject", we should not pass
through the private data, since it is most likely generated by the CM
without any protocol-specific knowledge.
Michael Brown [Mon, 10 Aug 2009 01:20:21 +0000 (02:20 +0100)]
[infiniband] Allow SRP reconnection attempts even after reporting failures
With iSCSI, connection attempts are expensive; it may take many
seconds to determine that a connection will fail. SRP connection
attempts are much less expensive, so we may as well avoid the
"optimisation" of declaring a state of permanent failure after a
certain number of attempts. This allows a gPXE SRP initiator to
resume operations after an arbitrary amount of SRP target downtime.
Michael Brown [Mon, 10 Aug 2009 10:32:13 +0000 (11:32 +0100)]
[infiniband] Generate more specific errors in response to failure MADs
Generate errors within individual MAD transaction consumers such as
ib_pathrec.c and ib_mcast.c, rather than within ib_mi.c. This allows
for more meaningful error messages to eventually be displayed to the
user.
Michael Brown [Fri, 17 Jul 2009 21:11:42 +0000 (22:11 +0100)]
[infiniband] Add support for SRP over Infiniband
SRP is the SCSI RDMA Protocol. It allows for a method of SAN booting
whereby the target is responsible for reading and writing data using
Remote DMA directly to the initiator's memory. The software initiator
merely sends and receives SCSI commands; it never has to touch the
actual data.
Michael Brown [Mon, 10 Aug 2009 10:47:11 +0000 (11:47 +0100)]
[infiniband] Add last_opened_ibdev(), analogous to last_opened_netdev()
The minimal-surprise behaviour, when no explicit SRP initiator device
is specified, will probably be to use the most recently opened
Infiniband device. This matches our behaviour with using the most
recently opened net device for PXE, iSCSI, AoE, NBI, etc.
Michael Brown [Sun, 9 Aug 2009 10:05:21 +0000 (11:05 +0100)]
[infiniband] Add a "communication-managed reliable connection" protocol
SRP over Infiniband uses a protocol whereby data is sent via a
combination of the CM private data fields and the RC queue pair
itself. This seems sufficiently generic that it's worth having
available as a separate protocol.
Michael Brown [Mon, 10 Aug 2009 01:10:17 +0000 (02:10 +0100)]
[hermon] Reduce the RC ACK timeout
The ACK timeout determines how long we take to notice a failed
Reliable Connection. Reducing it from the arbitrary value of 19 down
to 14 reduces the individual ACK timeout from around 2.1s to 67ms;
this in turn reduces the time to tear down and re-establish a broken
SRP session from around 30s to around 1s.
Michael Brown [Sun, 9 Aug 2009 19:38:35 +0000 (20:38 +0100)]
[hermon] Randomise the high-order bits of queue pair numbers
The Infiniband Communication Manager will refuse to establish a
connection if it believes the connection is already established.
There is no immediately obvious way to ask it to tear down the
existing connection and replace it; to issue a DREP we would need to
know the local and remote communication IDs used for the previous
connection setup.
We can work around this by randomising the high-order bits of the
queue pair number; these have no significance to the hardware, but are
sufficient to convince the IB CM that this is a different connection.
Michael Brown [Sun, 9 Aug 2009 00:25:38 +0000 (01:25 +0100)]
[infiniband] Handle duplicate Communication Management REPs
We will terminate our transaction as soon as we receive the first CM
REP, since that provides all the state that we need. However, the
peer may resend the REP if it didn't see our RTU, and if we don't
respond with another RTU we risk being disconnected. (This protocol
appears not to handle retries gracefully.)
Fix by adding a management agent that will listen for these duplicate
REPs and send back an RTU.
Joshua Oreman [Sat, 1 Aug 2009 16:56:44 +0000 (09:56 -0700)]
[802.11] Fix maximum packet length
Previously the maximum packet length was computed using an erroneous
understanding of the role of the MIC field in TKIP-encrypted packets.
The field is actually considered to be part of the MSDU (encrypted and
fragmented data), not the MPDU (container for each encrypted
fragment). As such its size does not contribute to cryptographic
overhead outside the data field's size limitations. The net result is
that the previous maximum packet length value was 4 bytes too long;
fix it to the correct value of 2352.
Signed-off-by: Michael Brown <mcb30@etherboot.org>
Joshua Oreman [Sat, 1 Aug 2009 16:52:59 +0000 (09:52 -0700)]
[802.11] Set channels early on to avoid tuning to an undefined channel
Some cards (such as ath5k) always need to tune to a particular channel
when they are reset; the reset may happen upon open(), which is before
the channels array would be set up (in prepare_probe()). Avoid tuning
the card to an inconsistent state by copying the hardware
supported-channels array to the 802.11 device's allowable-channels
array even before channels are "properly" set up.
Signed-off-by: Michael Brown <mcb30@etherboot.org>
Joshua Oreman [Fri, 7 Aug 2009 08:38:23 +0000 (01:38 -0700)]
[802.11] Enhance support for driver PHY differences
The prior net80211 model of physical-layer behavior for drivers was
overly simplistic and limited the drivers that could be written. To
be more flexible, split the driver-provided list of supported rates by
band, and add a means for specifying a list of supported channels.
Allow drivers to specify a hardware channel value that will be tied to
uses of the channel.
Expose net80211_duration() to drivers, and make the rate it uses in
its computations configurable, so that it can be used in calculating
durations that must be set in hardware for ACK and CTS packets. Add
net80211_cts_duration() for the common case of calculating the
duration for a CTS packet.
Signed-off-by: Michael Brown <mcb30@etherboot.org>
Michael Brown [Mon, 3 Aug 2009 14:56:56 +0000 (15:56 +0100)]
[infiniband] Add the concept of a management interface
A management interface is the component through which both local and
remote management agents are accessed.
This new implementation of a management interface allows for the user
to react to timed-out transactions, and also allows for cancellation
of in-progress transactions.
Michael Brown [Sat, 8 Aug 2009 13:36:10 +0000 (14:36 +0100)]
[romprefix] Cope with PnP BIOSes that fail to set %es:%di on entry
Some BIOSes support the BIOS Boot Specification (BBS) but fail to set
%es:%di correctly when calling the option ROM initialisation entry
point. This causes gPXE to identify the BIOS as non-PnP (and so
non-BBS), leaving the user unable to control the boot order.
Fix by scanning for the $PnP signature ourselves, rather than relying
on the BIOS having passed in %es:%di correctly.
Tested-by: Helmut Adrigan <helmut.adrigan@chello.at>
Michael Brown [Sun, 2 Aug 2009 21:57:01 +0000 (22:57 +0100)]
[infiniband] Change IB_{QPN,QKEY,QPT} names from {SMA,GMA} to {SMI,GSI}
The IBA specification refers to management "interfaces" and "agents".
The interface is the component that connects to the queue pair and
sends and receives MADs; the agent is the component that constructs
the reply to the MAD.
Rename the IB_{QPN,QKEY,QPT} constants as a first step towards making
this separation in gPXE.
Michael Brown [Mon, 3 Aug 2009 14:53:43 +0000 (15:53 +0100)]
[build] Mark __intel_new_proc_init with __libgcc rather than cdecl
The function __intel_new_proc_init() (called implicitly when building
using icc) is marked with __attribute__((cdecl)). This breaks
building on x86_64, where cdecl is meaningless.
Fix by replacing with the existing __libgcc macro, which is already
defined to be "__attribute__((cdecl))" for i386 builds and empty for
x86_64 builds.
H. Peter Anvin [Sun, 2 Aug 2009 21:41:38 +0000 (14:41 -0700)]
[pxe] Dual-license pxe_api.h under the MIT license
pxe_api.h is just a description of API functions, it's actively
undesirable to have more implementations than necessary. Allowing it
under the MIT license lets the Syslinux libraries use it.
Signed-off-by: H. Peter Anvin <hpa@zytor.com> Signed-off-by: Michael Brown <mcb30@etherboot.org>
Michael Brown [Sun, 2 Aug 2009 21:19:31 +0000 (22:19 +0100)]
[tcp] Avoid printf format warnings on some compilers
In several places, we currently use size_t to represent a difference
between TCP sequence numbers. This can cause compiler warnings
relating to printf format specifiers, since the result of
(uint32_t+size_t) may be an unsigned long on some compilers.
Fix by using uint32_t for all variables that represent a difference
between TCP sequence numbers.
Michael Brown [Sun, 2 Aug 2009 10:17:02 +0000 (11:17 +0100)]
[build] Allow safe concurrent builds of .iso, .liso and .sdsk targets
The geniso, genliso and gensdsk scripts contain hard-coded temporary
directory names, and so could potentially collide with each other when
run as part of a concurrent build (e.g. "make -j 4").
Fix by using mktemp to generate suitable temporary directory names.
Marty Connor [Sun, 29 Mar 2009 05:25:15 +0000 (01:25 -0400)]
[build] Add syslinux floppy image type .sdsk
We add a syslinux floppy disk type using parts of the genliso script.
This floppy image cat be dd'ed to a physical floppy or used in
instances where a virtual floppy with an mountable DOS filesystem is
useful.
We also modify the genliso script to only generate .liso images
rather than creating images depending on how it is called.
Signed-off-by: Michael Brown <mcb30@etherboot.org>
Michael Brown [Thu, 9 Jul 2009 04:18:33 +0000 (05:18 +0100)]
[hermon] Allow software GMA to receive packets destined for QP1
The Linux IB Communication Manager will always send MADs to QP1,
rather than back to the originating QP. On Hermon, QP1 is by default
handled by the embedded firmware. We can change this, but the cost is
that we have to handle both QP0 and QP1 (i.e. we have to provide SMA
as well as GMA service in software), and we have to use MLX queues
rather than standard UD queues (i.e. we have to construct the UD
datagrams by hand).
There doesn't seem to be any viable way around this situation, ugly
though it is.
Michael Brown [Fri, 17 Jul 2009 21:27:34 +0000 (22:27 +0100)]
[infiniband] Add infrastructure for RC queue pairs
Queue pairs are now assumed to be created in the INIT state, with a
call to ib_modify_qp() required to bring the queue pair to the RTS
state.
ib_modify_qp() no longer takes a modification list; callers should
modify the relevant queue pair parameters (e.g. qkey) directly and
then call ib_modify_qp() to synchronise the changes to the hardware.
The packet sequence number is now a property of the queue pair, rather
than of the device.
Each queue pair may have an associated address vector. For RC queue
pairs, this is the address vector that will be programmed in to the
hardware as the remote address. For UD queue pairs, it will be used
as the default address vector if none is supplied to ib_post_send().
Michael Brown [Fri, 10 Jul 2009 21:31:40 +0000 (22:31 +0100)]
[infiniband] Allow MAD handlers to indicate response via return value
Now that MAD handlers no longer return a status code, we can allow
them to return a pointer to a MAD structure if and only if they want
to send a response. This provides a more natural and flexible
approach than using a "response method" field within the handler's
descriptor.
Michael Brown [Fri, 10 Jul 2009 20:29:25 +0000 (21:29 +0100)]
[infiniband] Remove the return status code from MAD handlers
MAD handlers have to set the status fields within the MAD itself
anyway, in order to provide a meaningful response MAD; the additional
gPXE return status code is just noise.
Note that we probably don't need to ever explicitly set the status to
IB_MGMT_STATUS_OK, since it should already have this value from the
request. (By not explicitly setting the status in this way, we can
safely have ib_sma_set_xxx() call ib_sma_get_xxx() in order to
generate the GetResponse MAD without worrying that ib_sma_get_xxx()
will clear any error status set by ib_sma_set_xxx().)
Michael Brown [Thu, 9 Jul 2009 14:52:04 +0000 (15:52 +0100)]
[infiniband] Allow external QPN to differ from real QPN
Most IB hardware seems not to allow allocation of the genuine QPNs 0
and 1, so allow for the externally-visible QPN (as constructed and
parsed by ib_packet, where used) to differ from the real
hardware-allocated QPN.
Michael Brown [Tue, 7 Jul 2009 13:03:11 +0000 (14:03 +0100)]
[infiniband] Make qkey and rate optional parameters to ib_post_send()
The queue key is stored as a property of the queue pair, and so can
optionally be added by the Infiniband core at the time of calling
ib_post_send(), rather than always having to be specified by the
caller.
This allows IPoIB to avoid explicitly keeping track of the data queue
key.
Michael Brown [Tue, 7 Jul 2009 12:54:39 +0000 (13:54 +0100)]
[ipoib] Clarify new role of IPoIB peer cache as for MAC addresses only
Now that path record lookups are handled entirely via
ib_resolve_path(), the only role of the IPoIB peer cache is as a
lookup table for MAC addresses. Update the code structure and
comments to reflect this.
Michael Brown [Tue, 7 Jul 2009 12:30:47 +0000 (13:30 +0100)]
[ipoib] Expose the real broadcast MAC
The IPoIB broadcast MAC address varies according to the partition key.
Now that the broadcast MAC address is a property of the network device
rather than of the link layer, we can expose this real MAC address
directly.
The broadcast LID is now identified via a path record lookup; this is
marginally inefficient (since it was present in the MCMemberRecord
GetResponse), but avoids the need to special-case broadcasts when
constructing the address vector in ipoib_transmit().
Michael Brown [Mon, 6 Jul 2009 22:09:26 +0000 (23:09 +0100)]
[infiniband] Create a general management agent
Generalise the subnet management agent into a general management agent
capable of sending and responding to MADs, including support for
retransmissions as necessary.
Michael Brown [Mon, 6 Jul 2009 18:12:12 +0000 (19:12 +0100)]
[infiniband] Poll completion queues automatically
Currently, all Infiniband users must create a process for polling
their completion queues (or rely on a regular hook such as
netdev_poll() in ipoib.c).
Move instead to a model whereby the Infiniband core maintains a single
process calling ib_poll_eq(), and polling the event queue triggers
polls of the applicable completion queues. (At present, the
Infiniband core simply polls all of the device's completion queues.)
Polling a completion queue will now implicitly refill all attached
receive work queues; this is analogous to the way that netdev_poll()
implicitly refills the RX ring.
Infiniband users no longer need to create a process just to poll their
completion queues and refill their receive rings.
Michael Brown [Fri, 17 Jul 2009 21:50:33 +0000 (22:50 +0100)]
[infiniband] Centralise assumption of 2048-byte payloads
IPoIB and the SMA have separate constants for the packet size to be
used to I/O buffer allocations. Merge these into the single
IB_MAX_PAYLOAD_SIZE constant.
(Various other points in the Infiniband stack have hard-coded
assumptions of a 2048-byte payload; we don't currently support
variable MTUs.)
Michael Brown [Fri, 17 Jul 2009 21:48:31 +0000 (22:48 +0100)]
[netdevice] Make ll_broadcast per-netdevice rather than per-ll_protocol
IPoIB has a link-layer broadcast address that varies according to the
partition key. We currently go through several contortions to pretend
that the link-layer address is a fixed constant; by making the
broadcast address a property of the network device rather than the
link-layer protocol it will be possible to simplify IPoIB's broadcast
handling.
Michael Brown [Tue, 7 Jul 2009 22:01:28 +0000 (23:01 +0100)]
[ata] Make ATA command issuing partially asynchronous
Move the icky call to step() from aoe.c to ata.c; this takes it at
least one step further away from where it really doesn't belong.
Unfortunately, AoE has the ugly aoe_discover() mechanism which means
that we still have a step() loop in aoe.c for now; this needs to be
replaced at some future point.
Michael Brown [Mon, 6 Jul 2009 15:16:59 +0000 (16:16 +0100)]
[xfer] Always nullify interface while sending close() message
Objects typically call xfer_close() as part of their response to a
close() message. If the initiating object has already nullified the
xfer interface then this isn't a problem, but it can lead to
unexpected behaviour when the initiating object is aiming to reuse the
connection and so does not nullify the interface.
Fix by always temporarily nullifying the interface during xfer_close()
(as was already being done by xfer_vreopen() in order to work around
this specific problem).
Michael Brown [Sun, 28 Jun 2009 19:50:23 +0000 (20:50 +0100)]
[pxe] Add startpxe and stoppxe commands
These commands can be used to activate or deactivate the PXE API (on a
specifiable network interface).
This is currently of limited use, since most image formats will call
shutdown() before booting the image, meaning that the underlying net
device gets shut down during remove_devices() anyway.
Michael Brown [Sun, 28 Jun 2009 19:24:54 +0000 (20:24 +0100)]
[ifmgmt] Optimise prototype for ifcommon_exec()
ifcommon_exec() was long-ago marked as __attribute__((regparm(2))) in
order to minimise the size of functions that call into it. Since
then, gPXE has added -mregparm=3 as a general compilation option, and
this "optimisation" is now counter-productive.
Change (and simplify) the prototype to minimise code size given the
current compilation conditions.
Michael Brown [Sun, 28 Jun 2009 18:40:16 +0000 (19:40 +0100)]
[pxe] Make pxe_init_structures() an initialisation function
pxe_init_structures() fills in the fields of the !PXE and PXENV+
structures that aren't known until gPXE starts up. Once gPXE is
started, these values will never change.
Make pxe_init_structures() an initialisation function so that PXE
users don't have to worry about calling it.
Michael Brown [Sat, 27 Jun 2009 15:36:21 +0000 (16:36 +0100)]
[pxe] Update UNDI transmit count before transmitting packet
It is possible that the UNDI ISR may be triggered before netdev_tx()
returns control to pxenv_undi_transmit(). This means that
pxenv_undi_isr() may see a zero undi_tx_count, and so not check for TX
completions. This is not a significant problem, since it will check
for TX completions on the next call to pxenv_undi_isr() anyway; it
just means that the NBP will see a spurious IRQ that was apparently
caused by nothing.
Fix by updating the undi_tx_count before calling netdev_tx(), so that
pxenv_undi_isr() can decrement it and report the TX completion.
Symantec Ghost requires working multicast support. gPXE configures
all (sufficiently supported) network adapters into "receive all
multicasts" mode, which means that PXENV_UNDI_SET_MCAST_ADDRESS is
actually a no-op, but the current implementation returns
PXENV_STATUS_UNSUPPORTED instead.
Fix by making PXENV_UNDI_SET_MCAST_ADDRESS return success. For good
measure, also implement PXENV_UNDI_GET_MCAST_ADDRESS, since the
relevant functionality is now exposed by the net device core.
Note that this will silently fail if the gPXE driver for the NIC being
used fails to configure the NIC in "receive all multicasts" mode.
Michael Brown [Sat, 27 Jun 2009 13:43:10 +0000 (14:43 +0100)]
[pxe] Improve pxe_undi debug messages
The PXE debugging messages have remained pretty much unaltered since
Etherboot 5.4, and are now difficult to read in comparison to most of
the rest of gPXE.
Bring the pxe_undi debug messages up to normal gPXE standards.
Joshua Oreman [Fri, 19 Jun 2009 09:21:08 +0000 (02:21 -0700)]
[hci] Expose ifcommon_exec() in a local header so wireless commands can use it
This keeps code size down, since the wireless interface management
commands have the same command-line interface and overall structure as
the wired commands.
Signed-off-by: Michael Brown <mcb30@etherboot.org>
Joshua Oreman [Fri, 19 Jun 2009 09:08:21 +0000 (02:08 -0700)]
[ifmgmt] Move link-up status messages from autoboot() to iflinkwait()
With the addition of link status codes, we can now display a detailed
error indication if iflinkwait() fails.
Putting the error output in iflinkwait avoids code duplication, and
gains symmetry with the other interface management routines; ifopen()
already prints an error directly if it cannot open its interface.
Modified-by: Michael Brown <mcb30@etherboot.org> Signed-off-by: Michael Brown <mcb30@etherboot.org>
Michael Brown [Tue, 23 Jun 2009 21:53:29 +0000 (22:53 +0100)]
[pxe] Fix interoperability with the Symantec (undipd) DOS UNDI driver
The Symantec UNDI DOS driver fails when run on top of gPXE because we
return our interface type as "gPXE" rather than one of the predefined
NDIS interface type strings.
Fix by returning the standard "DIX+802.3" string; this isn't
necessarily always accurate, but it's highly unlikely that anything
trying to use the UNDI API would understand our IPoIB link-layer
pseudo-header anyway.