summaryrefslogtreecommitdiffstats
path: root/src
Commit message (Collapse)AuthorAgeFilesLines
* [usb] Add generic USB network device frameworkMichael Brown2015-02-173-0/+343
| | | | Signed-off-by: Michael Brown <mcb30@ipxe.org>
* [libc] Reduce size of memset()Michael Brown2015-02-171-1/+127
| | | | | | | | | As with memcpy(), we can reduce the code size (by an average of 0.2%) by giving the compiler more visibility into what memset() is doing, and by avoiding the "rep" prefix on short fixed-length sequences of string operations. Signed-off-by: Michael Brown <mcb30@ipxe.org>
* [test] Add constant-length memset() self-testsMichael Brown2015-02-172-0/+154
| | | | Signed-off-by: Michael Brown <mcb30@ipxe.org>
* [test] Add self-tests for more string functionsMichael Brown2015-02-171-0/+88
| | | | Signed-off-by: Michael Brown <mcb30@ipxe.org>
* [libc] Rewrite string functionsMichael Brown2015-02-176-403/+371Star
| | | | | | | | | | | | Some of the C library string functions have an unknown provenance. Reimplement all such functions to avoid potential licensing uncertainty. Remove the inline-assembler versions of strlen(), memswap(), and strncmp(); these save a minimal amount of space (around 40 bytes in total) and are not performance-critical. Signed-off-by: Michael Brown <mcb30@ipxe.org>
* [libc] Remove unused string functionsMichael Brown2015-02-162-194/+0Star
| | | | Signed-off-by: Michael Brown <mcb30@ipxe.org>
* [ncm] Use generic refill framework for bulk IN and interrupt endpointsMichael Brown2015-02-162-252/+92Star
| | | | Signed-off-by: Michael Brown <mcb30@ipxe.org>
* [ecm] Use generic refill framework for bulk IN and interrupt endpointsMichael Brown2015-02-132-133/+40Star
| | | | Signed-off-by: Michael Brown <mcb30@ipxe.org>
* [usb] Use generic refill framework for USB hub interrupt endpointsMichael Brown2015-02-132-51/+7Star
| | | | Signed-off-by: Michael Brown <mcb30@ipxe.org>
* [usb] Provide generic framework for refilling receive endpointsMichael Brown2015-02-132-0/+168
| | | | | | | Provide a generic framework for allocating, refilling, and optionally recycling I/O buffers used by bulk IN and interrupt endpoints. Signed-off-by: Michael Brown <mcb30@ipxe.org>
* [build] Apply the "-fno-PIE -nopie" workaround only to i386 buildsMichael Brown2015-02-122-11/+11
| | | | | | | | | | | | | | | | | | | | Hardened versions of gcc default to building position-independent code, which breaks our i386 build. Our build process therefore detects such platforms and automatically adds "-fno-PIE -nopie" to the gcc command line. On x86_64, we choose to build position-independent code (in order to reduce the final binary size and, in particular, the number of relocations required for UEFI binaries). The workaround therefore breaks the build process for x86_64 binaries on such platforms. Fix by moving the workaround to the i386-specific portion of the Makefile. Reported-by: Jan Kundrát <jkt@kde.org> Debugged-by: Jan Kundrát <jkt@kde.org> Debugged-by: Marin Hannache <git@mareo.fr> Signed-off-by: Michael Brown <mcb30@ipxe.org>
* [efi] Disallow R_X86_64_32 relocationsMichael Brown2015-02-121-2/+1Star
| | | | | | | | | | | | | | | | | UEFI binaries may be relocated to any location within the 64-bit address space. We compile as position-independent code with hidden visibility, which should force all relocation records to be either PC-relative (in which case no PE relocations are required) or full 64-bit relocations. There should be no R_X86_64_32 relocation records, since that would imply an invalid assumption that code could not be relocated above 4GB. Remove support for R_X86_64_32 relocation records from util/elf2efi.c, so that any such records result in a build failure rather than a potential runtime failure. Reported-by: Jan Kundrát <jkt@kde.org> Signed-off-by: Michael Brown <mcb30@ipxe.org>
* [build] Sort objects in blib.aOlaf Hering2015-02-121-1/+1
| | | | | | | | | | | | | When building hvmloader for Xen tools the iPXE objects are also linked into the binary. Unfortunately the linker will place them in the order found in the archive. Since this order is random the resulting hvmloader binary differs when it was built from identical sources but on different build hosts. To help with creating a reproducible binary the elements in blib.a must simply be sorted before passing them to $(AR). Signed-off-by: Olaf Hering <olaf@aepfle.de> Signed-off-by: Michael Brown <mcb30@ipxe.org>
* [usb] Handle port status changes received after failing to find a driverMichael Brown2015-02-122-6/+20
| | | | | | | | | | | | | | | | | | | | | Commit a60f2dd ("[usb] Try multiple USB device configurations") changed the behaviour of register_usb() such that if no drivers are found then the device will be closed and the memory used will be freed. If a port status change subsequently occurs while the device is still physically attached, then usb_hotplug() will see this as a new device having been attached, since there is no device recorded as being currently attached to the port. This can lead to spurious hotplug events (or even endless loops of hotplug events, if the process of opening and closing the device happens to generate a port status change). Fix by using a separate flag to indicate that a device is physically attached (even if we have no corresponding struct usb_device). Reported-by: Dan Ellis <Dan.Ellis@displaylink.com> Signed-off-by: Michael Brown <mcb30@ipxe.org>
* [rndis] Add rndis_rx_err()Michael Brown2015-02-112-0/+17
| | | | Signed-off-by: Michael Brown <mcb30@ipxe.org>
* [build] Allow product tag line to be customised via config/branding.hMichael Brown2015-02-112-1/+10
| | | | Signed-off-by: Michael Brown <mcb30@ipxe.org>
* [build] Allow setting help text URI to be customised via config/branding.hMichael Brown2015-02-113-3/+46
| | | | Signed-off-by: Michael Brown <mcb30@ipxe.org>
* [build] Allow command help text URI to be customised via config/branding.hMichael Brown2015-02-112-1/+46
| | | | Signed-off-by: Michael Brown <mcb30@ipxe.org>
* [build] Allow error message URI to be customised via config/branding.hMichael Brown2015-02-112-2/+48
| | | | Signed-off-by: Michael Brown <mcb30@ipxe.org>
* [build] Allow product URI to be customised via config/branding.hMichael Brown2015-02-113-3/+5
| | | | Signed-off-by: Michael Brown <mcb30@ipxe.org>
* [build] Use PRODUCT_SHORT_NAME for end-user visible stringsMichael Brown2015-02-114-6/+11
| | | | | | | | | | | Use PRODUCT_SHORT_NAME instead of a hardcoded "iPXE" for strings which are typically shown in the user interface. Note that this only allows for customisation of the user interface. Where the "iPXE" string serves a technical purpose (such as in the HTTP User-Agent), the string cannot be customised. Signed-off-by: Michael Brown <mcb30@ipxe.org>
* [build] Move branding information to config/branding.hMichael Brown2015-02-115-18/+36
| | | | Signed-off-by: Michael Brown <mcb30@ipxe.org>
* [xhci] Delay after (possibly) forcing port link state to RxDetectMichael Brown2015-02-112-0/+8
| | | | | | | | | | | | Some xHCI controllers (observed with a Renesas Electronics PCIe USB3 card) seem to require a delay after forcing the link state of USB3 ports to RxDetect. Omitting this delay causes strange behaviour including system lockups. Add an unconditional 20ms delay after writing the port link states. This seems to be sufficient to avoid the problem. Signed-off-by: Michael Brown <mcb30@ipxe.org>
* [ecm] Add support for CDC-ECM USB Ethernet devicesMichael Brown2015-02-102-1/+750
| | | | Signed-off-by: Michael Brown <mcb30@ipxe.org>
* [usb] Allow usb_stream() to enforce a terminating short packetMichael Brown2015-02-105-14/+28
| | | | | | | | | | Some USB endpoints require that a short packet be used to terminate transfers, since they have no other way to determine message boundaries. If the message length happens to be an exact multiple of the USB packet size, then this requires the use of an additional zero-length packet. Signed-off-by: Michael Brown <mcb30@ipxe.org>
* [usb] Parse endpoint descriptor bInterval fieldMichael Brown2015-02-104-5/+46
| | | | Signed-off-by: Michael Brown <mcb30@ipxe.org>
* [usb] Handle CDC union functional descriptorsMichael Brown2015-02-093-3/+100
| | | | | | | USB Communications Device Class devices may use a union functional descriptor to group several interfaces into a function. Signed-off-by: Michael Brown <mcb30@ipxe.org>
* [usb] Try multiple USB device configurationsMichael Brown2015-02-092-72/+137
| | | | | | | Iterate over a USB device's available configurations until we find one for which we have working drivers. Signed-off-by: Michael Brown <mcb30@ipxe.org>
* [ncm] Reserve headroom in received packetsMichael Brown2015-02-061-1/+11
| | | | | | | | | Some protocols (such as ARP) may modify the received packet and re-use the same I/O buffer for transmission of a reply. To allow this, reserve sufficient headroom at the start of each received packet buffer for our transmit datapath headers. Signed-off-by: Michael Brown <mcb30@ipxe.org>
* [pxe] Maintain a queue for received PXE UDP packetsMichael Brown2015-02-061-38/+85
| | | | | | | | | Some devices return multiple packets in a single poll. Handle such devices gracefully by enqueueing received PXE UDP packets (along with a pseudo-header to hold the IPv4 addresses and port numbers) and dequeueing them on subsequent calls to PXENV_UDP_READ. Signed-off-by: Michael Brown <mcb30@ipxe.org>
* [pxe] Use tftp_uri() to construct PXE TFTP URIsMichael Brown2015-02-061-12/+15
| | | | Signed-off-by: Michael Brown <mcb30@ipxe.org>
* [uri] Allow tftp_uri() to construct a URI with a custom portMichael Brown2015-02-064-7/+30
| | | | Signed-off-by: Michael Brown <mcb30@ipxe.org>
* [tftp] Explicitly abort connection whenever parent interface is closedMichael Brown2015-02-062-48/+23Star
| | | | | | | | | Fetching the TFTP file size is currently implemented via a custom "tftpsize://" protocol hack. Generalise this approach to instead close the TFTP connection whenever the parent data-transfer interface is closed. Signed-off-by: Michael Brown <mcb30@ipxe.org>
* [ncm] Use large multi-packet buffers by defaultMichael Brown2015-02-062-165/+301
| | | | | | | | | | | | | | | | | | | | | | | | Some devices have a very small number of internal buffers, and rely on being able to pack multiple packets into each buffer. Using 2048-byte buffers on such devices produces throughput of around 100Mbps. Using a small number of much larger buffers (e.g. 32kB) increases the throughput to around 780Mbps. (The full 1Gbps is not reached because the high RTT induced by the use of multi-packet buffers causes us to saturate our 256kB TCP window.) Since allocation of large buffers is very likely to fail, allocate the buffer set only once when the device is opened and recycle buffers immediately after use. Received data is now always copied to per-packet buffers. If allocation of large buffers fails, fall back to allocating a larger number of smaller buffers. This will give reduced performance, but the device will at least still be functional. Share code between the interrupt and bulk IN endpoint handlers, since the buffer handling is now very similar. Signed-off-by: Michael Brown <mcb30@ipxe.org>
* [usb] Report xHCI host controller eventsMichael Brown2015-02-052-0/+41
| | | | Signed-off-by: Michael Brown <mcb30@ipxe.org>
* [ncm] Add support for CDC-NCM USB Ethernet devicesMichael Brown2015-02-036-0/+1198
| | | | Signed-off-by: Michael Brown <mcb30@ipxe.org>
* [usb] Add support for xHCI host controllersMichael Brown2015-02-035-0/+4081
| | | | Signed-off-by: Michael Brown <mcb30@ipxe.org>
* [usb] Add basic support for USB hubsMichael Brown2015-02-035-0/+805
| | | | Signed-off-by: Michael Brown <mcb30@ipxe.org>
* [usb] Add basic support for USB devicesMichael Brown2015-02-035-1/+2822
| | | | Signed-off-by: Michael Brown <mcb30@ipxe.org>
* [romprefix] Ensure UNDI loader can be included by all ROM typesMichael Brown2015-02-023-0/+30
| | | | Signed-off-by: Michael Brown <mcb30@ipxe.org>
* [pci] Allow drivers to specify a PCI classMichael Brown2015-02-024-101/+63Star
| | | | | | | | Allow drivers to specify a supported PCI class code. To save space in the final binary, make this an attribute of the driver rather than an attribute of a PCI device ID list entry. Signed-off-by: Michael Brown <mcb30@ipxe.org>
* [build] Include Hyper-V driver in the all-drivers buildMichael Brown2014-12-211-0/+4
| | | | Signed-off-by: Michael Brown <mcb30@ipxe.org>
* [hyperv] Require support for VMBus version 3.0 or newerMichael Brown2014-12-212-17/+83
| | | | | | | | | | | | | | | | | | | | | | | | We require the ability to disconnect from and reconnect to VMBus; if we don't have this then there is no (viable) way for a loaded operating system to continue to use any VMBus devices. (There is also a small but non-zero risk that the host will continue to write to our interrupt and monitor pages, since the VMBUS_UNLOAD message in earlier versions is essentially a no-op.) This requires us to ensure that the host supports protocol version 3.0 (VMBUS_VERSION_WIN8_1). However, we can't actually _use_ protocol version 3.0, since doing so causes an iSCSI-booted Windows Server 2012 R2 VM to crash due to a NULL pointer dereference in vmbus.sys. To work around this problem, we first ensure that we can connect using protocol v3.0, then disconnect and reconnect using the oldest known protocol. This deliberately prevents the use of the iPXE native Hyper-V drivers on older versions of Hyper-V, where we could use our drivers but in so doing would break the loaded operating system. Signed-off-by: Michael Brown <mcb30@ipxe.org>
* [hyperv] Tidy up debug outputMichael Brown2014-12-201-2/+2
| | | | Signed-off-by: Michael Brown <mcb30@ipxe.org>
* [rndis] Ignore start-of-day RNDIS_INDICATE_STATUS_MSG with status 0x40020006Michael Brown2014-12-202-0/+6
| | | | | | | | | | | Windows Server 2012 R2 generates an RNDIS_INDICATE_STATUS_MSG with a status code of 0x4002006. This status code does not appear to be documented anywhere within the sphere of human knowledge. Explicitly ignore this status code in order to avoid unnecessarily cluttering the display when RNDIS debugging is enabled. Signed-off-by: Michael Brown <mcb30@ipxe.org>
* [hyperv] Assume that VMBus xfer page ranges correspond to RNDIS messagesMichael Brown2014-12-204-94/+72Star
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | The (undocumented) VMBus protocol seems to allow for transfer page-based packets where the data payload is split into an arbitrary set of ranges within the transfer page set. The RNDIS protocol includes a length field within the header of each message, and it is known from observation that multiple RNDIS messages can be concatenated into a single VMBus message. iPXE currently assumes that the transfer page range boundaries are entirely arbitrary, and uses the RNDIS header length to determine the RNDIS message boundaries. Windows Server 2012 R2 generates an RNDIS_INDICATE_STATUS_MSG for an undocumented and unknown status code (0x40020006) with a malformed RNDIS header length: the length does not cover the StatusBuffer portion of the message. This causes iPXE to report a malformed RNDIS message and to discard any further RNDIS messages within the same VMBus message. The Linux Hyper-V driver assumes that the transfer page range boundaries correspond to RNDIS message boundaries, and so does not notice the malformed length field in the RNDIS header. Match the behaviour of the Linux Hyper-V driver: assume that the transfer page range boundaries correspond to the RNDIS message boundaries and ignore the RNDIS header length. This avoids triggering the "malformed packet" error and also avoids unnecessary data copying: since we now have one I/O buffer per RNDIS message, there is no longer any need to use iob_split(). Signed-off-by: Michael Brown <mcb30@ipxe.org>
* [hyperv] Increase TX ring sizeMichael Brown2014-12-201-1/+1
| | | | | | | | Empirical observation suggests that 32 is a sensible size to minimise the number of deferred packet transmissions without overflowing the VMBus transmit ring buffer. Signed-off-by: Michael Brown <mcb30@ipxe.org>
* [hyperv] Receive all VMBus messages in a pollMichael Brown2014-12-201-1/+3
| | | | | | | Allow for elision of transmitted TCP ACKs by handling all received VMBus messages in each network device poll operation. Signed-off-by: Michael Brown <mcb30@ipxe.org>
* [rndis] Clear receive filter when closing the deviceMichael Brown2014-12-201-11/+30
| | | | | | | | | On Windows Server 2012 R2, closing and reopening the device will sometimes result in a non-functional RX datapath. The root cause is unknown. Clearing the receive filter before closing the device seems to fix the problem. Signed-off-by: Michael Brown <mcb30@ipxe.org>
* [hyperv] Tear down NetVSC RX buffer GPADL after closing VMBus deviceMichael Brown2014-12-201-24/+20Star
| | | | | | | | | | | | | On Windows Server 2012 R2, the receive buffer teardown completion message seems to occasionally be deferred until after the VMBus channel has been closed. This happens even if there are no packets currently in the receive buffer. Work around this problem by separating the revocation and teardown of the receive buffer, and deferring the teardown until after the VMBus channel has been closed. Signed-off-by: Michael Brown <mcb30@ipxe.org>
generated by cgit v1.2.3-55-g7522 (git 2.39.0) at 2024-07-12 14:20:08 +0200