aboutsummaryrefslogtreecommitdiffstats
path: root/bitbang_spi.c
Commit message (Collapse)AuthorAgeFilesLines
* bitbang_spi: Drop bitbang_spi_master_typeNico Huber2019-06-231-6/+2
| | | | | | | | | | | It only existed to make maintenance harder and waste our time. Change-Id: I7a3b5d9ff1e99d2d4f873c6f19fb318f93762037 Signed-off-by: Nico Huber <nico.h@gmx.de> Reviewed-on: https://review.coreboot.org/c/flashrom/+/33638 Tested-by: build bot (Jenkins) <no-reply@coreboot.org> Reviewed-by: HAOUAS Elyes <ehaouas@noos.fr> Reviewed-by: Angel Pons <th3fanbus@gmail.com>
* bitbang_spi: Add half-duplex optimizationsDaniel Thompson2018-08-171-5/+19
| | | | | | | | | | | | | | | | | | Currently, the core of bitbang_spi is a full-duplex SPI loop but in practice this code is only ever used half-duplex. Spliting this code into two half duplex loops allows us to optimize performance by reducing communications and/or CPU pipeline stalls. The speed up varies depending on how much the overhead of getting/setting pins dominates execution time. For a USB bit bang driver running on a 7th generation Core i5, the time to probe drops from ~7.7 seconds to ~6.7 seconds when this patch is applied. Change-Id: I33b9f363716f651146c09113bda5fffe53b16738 Signed-off-by: Daniel Thompson <daniel.thompson@linaro.org> Reviewed-on: https://review.coreboot.org/26947 Tested-by: build bot (Jenkins) <no-reply@coreboot.org> Reviewed-by: Nico Huber <nico.h@gmx.de>
* bitbang_spi: Add functions to optimize xfersDaniel Thompson2018-08-171-16/+24
| | | | | | | | | | | | | | | | | | | | On systems where the overhead of getting/setting pins is much greater than the half period (for example, USB bit banging) it significantly boosts performance if we can bang more than one bit at the same time. Add support for setting sck at the same time as mosi or miso activity. The speed up varies depending on how much the overhead of getting/setting pins dominates execution time. For a USB bit bang driver running on a 7th generation Core i5, the time to probe drops from ~9.2 seconds to ~7.7 seconds when set_clk_set_mosi() is implemented. Change-Id: Ic3430a9df34844cdfa82e109456be788eaa1789a Signed-off-by: Daniel Thompson <daniel.thompson@linaro.org> Reviewed-on: https://review.coreboot.org/26946 Tested-by: build bot (Jenkins) <no-reply@coreboot.org> Reviewed-by: Patrick Georgi <pgeorgi@google.com> Reviewed-by: Idwer Vollering <vidwer@gmail.com> Reviewed-by: Nico Huber <nico.h@gmx.de>
* Remove address from GPLv2 headersElyes HAOUAS2018-04-241-4/+0
| | | | | | | | Change-Id: I7bfc339673cbf5ee2d2ff7564c4db04ca088d0a4 Signed-off-by: Elyes HAOUAS <ehaouas@noos.fr> Reviewed-on: https://review.coreboot.org/25381 Tested-by: build bot (Jenkins) <no-reply@coreboot.org> Reviewed-by: Nico Huber <nico.h@gmx.de>
* spi_master: Introduce SPI_MASTER_4BA feature flagNico Huber2018-01-021-0/+1
| | | | | | | | | | | | | | | Add a feature flag SPI_MASTER_4BA to `struct spi_master` that advertises programmer-side support for 4-byte addresses in generic commands (and read/write commands if the master uses the default implementations). Set it for all masters that handle commands address-agnostic. Don't prefer native 4BA instructions if the master doesn't support them. Change-Id: Ife66e3fc49b9716f9c99cad957095b528135ec2c Signed-off-by: Nico Huber <nico.h@gmx.de> Reviewed-on: https://review.coreboot.org/22421 Tested-by: build bot (Jenkins) <no-reply@coreboot.org> Reviewed-by: David Hendricks <david.hendricks@gmail.com>
* Rename programmer registration functionsCarl-Daniel Hailfinger2014-07-191-6/+6
| | | | | | | | | | | | Register_programmer suggests that we register a programmer. However, that function registers a master for a given bus type, and a programmer may support multiple masters (e.g. SPI, FWH). Rename a few other functions to be more consistent. Corresponding to flashrom svn r1831. Signed-off-by: Carl-Daniel Hailfinger <c-d.hailfinger.devel.2006@gmx.net> Acked-by: Stefan Tauner <stefan.tauner@alumni.tuwien.ac.at>
* Fix duplicate 'const' declaration specifiersStefan Tauner2013-01-151-6/+6
| | | | | | | | | | Thanks to Idwer and clang for noticing these problems. Corresponding to flashrom svn r1646. Signed-off-by: Stefan Tauner <stefan.tauner@alumni.tuwien.ac.at> Signed-off-by: Idwer Vollering <vidwer@gmail.com> Acked-by: Stefan Tauner <stefan.tauner@alumni.tuwien.ac.at>
* Let the programmer driver decide how to do AAI transfersNico Huber2012-06-151-0/+1
| | | | | | | | | | | | | | | | Currently spi_aai_write() is implemented without an abstraction mechanism for the programmer driver. This adds another function pointer 'write_aai' to struct spi_programmer, which is set to default_spi_write_aai (renamed spi_aai_write) for all programmers for now. A patch which utilises this abstraction in the dediprog driver will follow. Corresponding to flashrom svn r1543. Signed-off-by: Nico Huber <nico.huber@secunet.com> Acked-by: Carl-Daniel Hailfinger <c-d.hailfinger.devel.2006@gmx.net>
* Have all programmer init functions register bus masters/programmersCarl-Daniel Hailfinger2011-12-201-68/+56
| | | | | | | | | | | | | | | | | | | | | | | | | | | | All programmer types (Parallel, SPI, Opaque) now register themselves into a generic programmer list and probing is now programmer-centric instead of chip-centric. Registering multiple SPI/... masters at the same time is now possible without any problems. Handling multiple flash chips is still unchanged, but now we have the infrastructure to deal with "dual BIOS" and "one flash behind southbridge and one flash behind EC" sanely. A nice side effect is that this patch kills quite a few global variables and improves the situation for libflashrom. Hint for developers: struct {spi,par,opaque}_programmer now have a void *data pointer to store any additional programmer-specific data, e.g. hardware configuration info. Note: flashrom -f -c FOO -r forced_read.bin does not work anymore. We have to find an architecturally clean way to solve this. Corresponding to flashrom svn r1475. Signed-off-by: Carl-Daniel Hailfinger <c-d.hailfinger.devel.2006@gmx.net> Acked-by: Michael Karcher <flashrom@mkarcher.dialup.fu-berlin.de>
* Add struct flashctx * parameter to all functions accessing flash chipsCarl-Daniel Hailfinger2011-12-181-4/+8
| | | | | | | | | | | | | | | | All programmer access function prototypes except init have been made static and moved to the respective file. A few internal functions in flash chip drivers had chipaddr parameters which are no longer needed. The lines touched by flashctx changes have been adjusted to 80 columns except in header files. Corresponding to flashrom svn r1474. Signed-off-by: Carl-Daniel Hailfinger <c-d.hailfinger.devel.2006@gmx.net> Acked-by: Michael Karcher <flashrom@mkarcher.dialup.fu-berlin.de>
* Remove unneeded inclusions of chipdrivers.hStefan Tauner2011-08-161-1/+0
| | | | | | | | | | | | | | | | | | | | | | This is related to the spi split patch as discussed in: http://www.flashrom.org/pipermail/flashrom/2010-February/thread.html#2364 the old commit (r914) log notes: "Some of the spi programmer drivers required chipdrivers.h, needs fixing later: it87spi.c ichspi.c sb600spi.c wbsio_spi.c buspirate_spi.c ft2232spi.c bitbang_spi.c dediprog.c" there still remain a few cases where chipdrivers.h is needed: dediprog.c (spi_read_chunked and spi_write_chunked) it87spi.c (due to spi_write_enable and spi_read_status_register) wbsio_spi.c (spi_programmer registration only) besides that, there are also non-spi files that do not need it. also, add flash.h to chipdrivers.h because it uses some types of it and remove flashchips.h from print.c Corresponding to flashrom svn r1414. Signed-off-by: Stefan Tauner <stefan.tauner@alumni.tuwien.ac.at> Acked-by: Stefan Tauner <stefan.tauner@alumni.tuwien.ac.at>
* Random whitespace and coding-style fixesUwe Hermann2011-07-281-7/+7
| | | | | | | | | | | | Also, indentation fixes, e.g. due to conversion to msg_*, use ARRAY_SIZE where possible, wrap overly long line, etc. Compile-tested. There should be no functional changes. Corresponding to flashrom svn r1397. Signed-off-by: Uwe Hermann <uwe@hermann-uwe.de> Acked-by: Uwe Hermann <uwe@hermann-uwe.de>
* Kill central list of SPI programmersMichael Karcher2011-05-111-1/+16
| | | | | | | | | | | | | | | | | | Remove the array spi_programmer, replace it by dynamic registration instead. Also initially start with no busses supported, and switch to the default non-SPI only for the internal programmer. Also this patch changes the initialization for the buses_supported variable from "everything-except-SPI" to "nothing". All programmers have to set the bus type on their own, and this enables register_spi_programmer to just add the SPI both for on-board SPI interfaces (where the internal programmer already detected the other bus types), as well as for external programmers (where we have the default "none"). Corresponding to flashrom svn r1299. Signed-off-by: Michael Karcher <flashrom@mkarcher.dialup.fu-berlin.de> Acked-by: Carl-Daniel Hailfinger <c-d.hailfinger.devel.2006@gmx.net>
* Factor out SPI write/read chunking wrappersMichael Karcher2011-05-111-11/+0
| | | | | | | Corresponding to flashrom svn r1298. Signed-off-by: Michael Karcher <flashrom@mkarcher.dialup.fu-berlin.de> Acked-by: Carl-Daniel Hailfinger <c-d.hailfinger.devel.2006@gmx.net>
* Remove duplicate includes from the codeStefan Reinauer2010-10-061-1/+0
| | | | | | | Corresponding to flashrom svn r1196. Signed-off-by: Stefan Reinauer <stepan@coresystems.de> Acked-by: Carl-Daniel Hailfinger <c-d.hailfinger.devel.2006@gmx.net>
* SPI bitbanging: request/release busCarl-Daniel Hailfinger2010-09-151-2/+46
| | | | | | | | | | | | | | | | | | | | | | | | | SPI bitbanging on devices which speak SPI natively has a dual-use problem: We need to shut down normal SPI operations to do the bitbanging ourselves. Once we're done, it makes a lot of sense to reenable "normal" SPI operations again. Add request_bus/release_bus functions to struct bitbang_spi_master. Add a bitbang shutdown function (not used yet). Change MCP SPI and Intel NIC SPI to use the new request/release bus infrastructure. Cosmetic changes to a few error messages (80 column limit). There are multiple possible strategies for bus request/release: - Request at the start of a SPI command, release immediately afterwards. - Request at the start of a SPI multicommand, release once all commands of the multicommand are done. - Request on programmer init, release on shutdown. Each strategy has its own advantages. For now, we will stay with the first strategy which worked fine so far. Corresponding to flashrom svn r1171. Signed-off-by: Carl-Daniel Hailfinger <c-d.hailfinger.devel.2006@gmx.net> Acked-by: Uwe Hermann <uwe@hermann-uwe.de>
* Split off programmer.h from flash.hCarl-Daniel Hailfinger2010-07-271-0/+1
| | | | | | | | | | | | | | | | | | | | | Programmer specific functions are of absolutely no interest to any file except those dealing with programmer specific actions (special SPI commands and the generic core). The new header structure is as follows (and yes, improvements are possible): flashchips.h flash chip IDs chipdrivers.h chip-specific read/write/... functions flash.h common header for all stuff that doesn't fit elsewhere hwaccess.h hardware access functions programmer.h programmer specific functions coreboot_tables.h header from coreboot, internal programmer only spi.h SPI command definitions Corresponding to flashrom svn r1112. Signed-off-by: Carl-Daniel Hailfinger <c-d.hailfinger.devel.2006@gmx.net> Acked-by: Uwe Hermann <uwe@hermann-uwe.de>
* Check during SPI bitbang init that SPI bitbang master configuration is correctCarl-Daniel Hailfinger2010-07-181-0/+10
| | | | | | | Corresponding to flashrom svn r1092. Signed-off-by: Carl-Daniel Hailfinger <c-d.hailfinger.devel.2006@gmx.net> Acked-by: Michael Karcher <flashrom@mkarcher.dialup.fu-berlin.de>
* Use struct pointer instead of enum to set bitbang adapterMichael Karcher2010-07-171-17/+6
| | | | | | | Corresponding to flashrom svn r1091. Signed-off-by: Michael Karcher <flashrom@mkarcher.dialup.fu-berlin.de> Acked-by: Carl-Daniel Hailfinger <c-d.hailfinger.devel.2006@gmx.net>
* Refine SPI bitbangingCarl-Daniel Hailfinger2010-07-171-11/+20
| | | | | | | | | | | | | | | Change the SPI bitbanging core to fix a subtle bug (which had no effect so far) and to make integration of the RayeR SPIPGM and Nvidia MCP6x/MCP7x SPI patches easier. Kill a few global variables and require explicit initialization of bitbanging delay. A big to Johannes Sjölund for testing an earlier version of the code as part of the Nvidia MCP6x/MCP7x SPI bitbanging patch. Corresponding to flashrom svn r1085. Signed-off-by: Carl-Daniel Hailfinger <c-d.hailfinger.devel.2006@gmx.net> Acked-by: Michael Karcher <flashrom@mkarcher.dialup.fu-berlin.de>
* Remove temporary buffers from bitbangingMichael Karcher2010-07-171-38/+5
| | | | | | | | | | | | | This removes the need of allocating an extra buffer, but also removes the possibility of having the data read back during the initial write phase for debugging purposes. Compile tested, no functional testing. Corresponding to flashrom svn r1084. Signed-off-by: Michael Karcher <flashrom@mkarcher.dialup.fu-berlin.de> Acked-by: Carl-Daniel Hailfinger <c-d.hailfinger.devel.2006@gmx.net>
* Use generic unlocking infrastructure for SPI chipsCarl-Daniel Hailfinger2010-07-141-1/+0
| | | | | | | | | Actually check if the unlock worked instead of just assuming it worked. Corresponding to flashrom svn r1082. Signed-off-by: Carl-Daniel Hailfinger <c-d.hailfinger.devel.2006@gmx.net> Acked-by: Michael Karcher <flashrom@mkarcher.dialup.fu-berlin.de>
* Convert SPI chips to partial writeCarl-Daniel Hailfinger2010-07-141-5/+3
| | | | | | | | | | However, wrap the write functions in a compat layer to allow converting the rest of flashrom later. Tested on Intel NM10 by David Hendricks. Corresponding to flashrom svn r1080. Signed-off-by: Carl-Daniel Hailfinger <c-d.hailfinger.devel.2006@gmx.net> Acked-by: Michael Karcher <flashrom@mkarcher.dialup.fu-berlin.de>
* Introduce a generic SPI read function: spi_write_chunked()Carl-Daniel Hailfinger2010-05-211-19/+2
| | | | | | | | | | | | | | | Every SPI programmer driver had its own completely different chip write implementation, and all of them were insufficiently commented. Create spi_write_chunked as a copy of spi_read_chunked and convert all SPI programmers to use it. No functional changes except: - Bus Pirate uses 12 Byte writes instead of 8 Byte writes - SB600 uses 5 Byte writes instead of 1 Byte writes Corresponding to flashrom svn r1005. Signed-off-by: Carl-Daniel Hailfinger <c-d.hailfinger.devel.2006@gmx.net> Acked-by: Michael Karcher <flashrom@mkarcher.dialup.fu-berlin.de> Acked-by: David Hendricks <dhendrix@google.com>
* Split spi.c into programmer and chip code Remove chipdriver.h include from ↵Sean Nelson2010-02-261-0/+1
| | | | | | | | | | | | | | | | | | | flash.h Some of the spi programmer drivers required chipdrivers.h, needs fixing later: it87spi.c ichspi.c sb600spi.c wbsio_spi.c buspirate_spi.c ft2232spi.c bitbang_spi.c dediprog.c Corresponding to flashrom svn r914. Signed-off-by: Sean Nelson <audiohacked@gmail.com> Acked-by: Carl-Daniel Hailfinger <c-d.hailfinger.devel.2006@gmx.net>
* Convert all messages in bitbang_spi.c to the new message infrastructureSean Nelson2010-01-091-4/+4
| | | | | | | Corresponding to flashrom svn r852. Signed-off-by: Sean Nelson <audiohacked@gmail.com> Acked-by: Carl-Daniel Hailfinger <c-d.hailfinger.devel.2006@gmx.net>
* Reduce realloc syscall overhead for FT2232 and bitbangCarl-Daniel Hailfinger2009-11-251-2/+5
| | | | | | | | | | | | FT2232 ran realloc() for every executed command. Start with a big enough buffer and don't touch buffer size unless it needs to grow. Bitbang was slightly better: It only ran realloc() if buffer size changed. Still, the solution above improves performance and reliability. Corresponding to flashrom svn r780. Signed-off-by: Carl-Daniel Hailfinger <c-d.hailfinger.devel.2006@gmx.net> Acked-by: Sean Nelson <audiohacked@gmail.com>
* Make bitbang_spi naming consistentCarl-Daniel Hailfinger2009-10-011-26/+26
| | | | | | | Corresponding to flashrom svn r741. Signed-off-by: Carl-Daniel Hailfinger <c-d.hailfinger.devel.2006@gmx.net> Acked-by: Uwe Hermann <uwe@hermann-uwe.de>
* This is the bitbanging SPI driver infrastructureCarl-Daniel Hailfinger2009-09-281-0/+161
If you want support for a particular piece of hardware, just fill in a few functions in spi_bitbang_master_table. That's it. On top of this, the RayeR SPI flasher should be supportable in ~20 LOC. Tested, trace looks OK. Corresponding to flashrom svn r736. Signed-off-by: Carl-Daniel Hailfinger <c-d.hailfinger.devel.2006@gmx.net> Acked-by: Uwe Hermann <uwe@hermann-uwe.de>