Carl-Daniel Hailfinger has uploaded this change for review. ( https://review.coreboot.org/c/flashrom/+/38488 )
Change subject: Automatic programmer driver writer ......................................................................
Automatic programmer driver writer
Edit build_new_driver.sh and fill in the properties of the flashrom programmer driver you want to create. Then run ./build_new_driver.sh
The driver skeleton will be created in $PROGRAMMERNAME.c.mine Modified versions of existing files will be created with extension .mine You can replace the original files with the modified versions by running for a in *; do test -f $a.mine && mv $a.mine $a; done
If you want to use the newly generated skeleton $PROGRAMMERNAME.c.mine , run mv $PROGRAMMERNAME.c.mine $PROGRAMMERNAME.c
WARNING: Please note that rerunning build_new_driver.sh will overwrite all *.mine files, but it won't touch $PROGRAMMERNAME.c . If something goes wrong, you can revert all files which look odd and run this script again.
TODO: Stefan Tauner suggested to put the configuration into a separate file
Change-Id: Icdcdcdcdcdcdcdcdcdcdcdcdcdcdcdcdcd000001 Signed-off-by: Carl-Daniel Hailfinger c-d.hailfinger.devel.2006@gmx.net --- M Makefile A build_new_driver.sh M flashrom.8.tmpl M flashrom.c M meson.build M meson_options.txt M programmer.h 7 files changed, 443 insertions(+), 1 deletion(-)
git pull ssh://review.coreboot.org:29418/flashrom refs/changes/88/38488/1
diff --git a/Makefile b/Makefile index 7242b09..3149bb6 100644 --- a/Makefile +++ b/Makefile @@ -696,6 +696,8 @@ # Disable J-Link for now. CONFIG_JLINK_SPI ?= no
+#PLACEHOLDER_NEWPROG_DEFAULTCONFIG + # Disable wiki printing by default. It is only useful if you have wiki access. CONFIG_PRINT_WIKI ?= no
@@ -759,7 +761,9 @@ ifeq ($(CONFIG_OGP_SPI), yes) override CONFIG_BITBANG_SPI = yes else +#PLACEHOLDER_NEWPROG_BITBANGSPICONFIG1 CONFIG_BITBANG_SPI ?= no +#PLACEHOLDER_NEWPROG_BITBANGSPICONFIG2 endif endif endif @@ -996,6 +1000,8 @@ PROGRAMMER_OBJS += mstarddc_spi.o endif
+#PLACEHOLDER_NEWPROG_COMPILERULE + ifeq ($(CONFIG_CH341A_SPI), yes) FEATURE_CFLAGS += -D'CONFIG_CH341A_SPI=1' PROGRAMMER_OBJS += ch341a_spi.o diff --git a/build_new_driver.sh b/build_new_driver.sh new file mode 100644 index 0000000..d15e135 --- /dev/null +++ b/build_new_driver.sh @@ -0,0 +1,420 @@ +#!/bin/bash +# flashrom programmer driver skeleton builder. +# Copyright 2012,2020 Carl-Daniel Hailfinger +# Licensed under the GNU GPL v2 +# The license of the generated programmer driver is unrelated to the license +# of this script and can be specified below. + +# Fill in all info in the block below, and don't touch anything else. +# The data provided here is just an example. +# Name of the programmer. Needs to be an all-lowercase valid C identifier. +PROGRAMMERNAME=ezo +# Short description of the programmer. Please do not use / inside the name, it will break the sed expressions. +PROGRAMMERDESCR="EZo+Willem Programmer" +# Name of the programmer manufacturer. +PROGRAMMERMANUF="EZo and Willem" +# Website for the programmer. +PROGRAMMERURL="http://www.ezoflash.com/" +# Fill in your name here. +AUTHORNAME="Carl-Daniel Hailfinger" +# License version of the new programmer driver: 2 or 2+ (for 2+later) +LICENSE_GPL=2 +# Does the programmer need a map/unmap function? +HAVE_MAP=no +# Does the programmer have its own delay function? +HAVE_DELAY=no +# Does the programmer need some sort of direct hardware access? +NEED_PCI=yes +# Does the programmer need some sort of serial port access? +NEED_SERIAL=no +# Is the programmer a PCI device, USB device, or something else? +# You have to specify exactly one of PCI, USB, OTHER +DEVICETYPE=USB +# Note: Usually a programmer only has one of NEED_PARLPCFWH, NEED_SPI or NEED_SPI_BITBANG set to yes. +# Does the programmer use Parallel/LPC/FWH functionality? +NEED_PARLPCFWH=yes +# Which of PARALLEL/LPC/FWH buses does the programer use? FIXME: Explain how to handle multiple buses. +BUS_PARLPCFWH=LPC +# Does the programmer use SPI functionality without bitbanging? FIXME: Check if a SPI bitbanging driver with NEED_SPI=no generates useful code. +NEED_SPI=yes +# Does the programmer use the bitbanging SPI infrastructure? +NEED_SPI_BITBANG=yes + +# No user serviceable parts below. +unset LANG +unset LANGUAGE +unset LC_COLLATE +if test $LICENSE_GPL = 2; then + GPLV3EITHER= + GPLV3OR= +elif test $LICENSE_GPL = 2+; then + GPLV3EITHER="either" + GPLV3OR="\n * (at your option) any later version" +else + echo "Specified license can not be handled automatically" + exit 1 +fi +if test $HAVE_MAP = yes; then MAPNAME=$PROGRAMMERNAME; else MAPNAME=fallback; fi +if test $HAVE_DELAY = yes; then DELAYNAME=$PROGRAMMERNAME; else DELAYNAME=internal; fi +PROGRAMMERNAMECAPS=$(echo -n $PROGRAMMERNAME|tr "[[:lower:]]" "[[:upper:]]") +CONFIGNAME=CONFIG_$PROGRAMMERNAMECAPS +ENUMNAME=PROGRAMMER_$PROGRAMMERNAMECAPS +if test $NEED_PCI = yes; then NEEDS="NEED_PCI := yes\n"; fi +if test $NEED_SERIAL = yes; then NEEDS+="NEED_SERIAL := yes\n"; fi + +sed "s-^//PLACEHOLDER_NEWPROG_PROGRAMMER_ARRAY-\ +#if ${CONFIGNAME} == 1\n\ + {\n\ + .name = "${PROGRAMMERNAME}",\n\ +\0-" flashrom.c >flashrom.c.mine +if test $DEVICETYPE = OTHER; then +sed "s-^//PLACEHOLDER_NEWPROG_PROGRAMMER_ARRAY-\ + .type = OTHER,\n\ + .devs.note = "Textual list of usable devices\\n",\n\ +\0-" flashrom.c.mine >flashrom.c.mine1 +mv flashrom.c.mine1 flashrom.c.mine +else +sed "s-^//PLACEHOLDER_NEWPROG_PROGRAMMER_ARRAY-\ + .type = ${DEVICETYPE},\n\ + .devs.dev = devs_${PROGRAMMERNAME},\n\ +\0-" flashrom.c.mine >flashrom.c.mine1 +mv flashrom.c.mine1 flashrom.c.mine +fi +sed "s-^//PLACEHOLDER_NEWPROG_PROGRAMMER_ARRAY-\ + .init = ${PROGRAMMERNAME}_init,\n\ + .map_flash_region = ${MAPNAME}_map,\n\ + .unmap_flash_region = ${MAPNAME}_unmap,\n\ + .delay = ${DELAYNAME}_delay,\n\ + },\n\ +#endif\n\ +\n\0-" flashrom.c.mine >flashrom.c.mine1 +mv flashrom.c.mine1 flashrom.c.mine + +sed -e "s/^#PLACEHOLDER_NEWPROG_DEFAULTCONFIG/\ +# Enable ${PROGRAMMERDESCR} for now.\n\ +${CONFIGNAME} ?= yes\n\ +\n\0/" \ +-e "s/^#PLACEHOLDER_NEWPROG_COMPILERULE/\ +ifeq ($(${CONFIGNAME}), yes)\n\ +FEATURE_CFLAGS += -D'${CONFIGNAME}=1'\n\ +PROGRAMMER_OBJS += ${PROGRAMMERNAME}.o\n\ +${NEEDS}\ +endif\n\ +\n\0/" Makefile >Makefile.mine + +if test $NEED_SPI_BITBANG = yes; then +sed -e "s/^#PLACEHOLDER_NEWPROG_BITBANGSPICONFIG1/\ +ifeq ($(${CONFIGNAME}), yes)\n\ +override CONFIG_BITBANG_SPI = yes\n\ +else\n\ +\0/" \ +-e "s/^#PLACEHOLDER_NEWPROG_BITBANGSPICONFIG2/\ +\0\n\ +endif/;" Makefile.mine >Makefile.mine1 +mv Makefile.mine1 Makefile.mine +fi + +sed -e "s-^//PLACEHOLDER_NEWPROG_PROGRAMMER_ENUM-\ +#if ${CONFIGNAME} == 1\n\ + ${ENUMNAME},\n\ +#endif\n\ +\0-" \ +-e "s-^//PLACEHOLDER_NEWPROG_PUBLICFUNCTIONS-\ +/* ${PROGRAMMERNAME}.c */\n\ +#if ${CONFIGNAME} == 1\n\ +int ${PROGRAMMERNAME}_init(void);\n\ +\0-" programmer.h >programmer.h.mine + +if test $DEVICETYPE = PCI -o $DEVICETYPE = USB; then +sed -e "s-^//PLACEHOLDER_NEWPROG_PUBLICFUNCTIONS-\ +extern const struct dev_entry devs_${PROGRAMMERNAME}[];\n\ +\n\0-" programmer.h.mine >programmer.h.mine1 +mv programmer.h.mine1 programmer.h.mine +fi + +sed -e "s-^//PLACEHOLDER_NEWPROG_PUBLICFUNCTIONS-\ +#endif\n\ +\n\0-" programmer.h.mine >programmer.h.mine1 +mv programmer.h.mine1 programmer.h.mine + +if test $NEED_SPI_BITBANG = yes; then +sed -e "s-//PLACEHOLDER_NEWPROG_SELECT_SPI_BITBANG$-\ +|| ${CONFIGNAME} == 1 \0-" programmer.h.mine >programmer.h.mine1 +mv programmer.h.mine1 programmer.h.mine +fi + +# No idea if roff supports hidden comments. Hook up to hopefully unchanged sequences. +sed -e "s/.*PLACEHOLDER_NEWPROG_MAN_SHORTDESCRIPTION/\ +.BR "* ${PROGRAMMERNAME}" " (${PROGRAMMERDESCR})"\n\ +.sp\n\ +\0/" \ +-e "s/.*PLACEHOLDER_NEWPROG_MAN_LONGDESCRIPTION/\ +.SS\n\ +.BR "${PROGRAMMERNAME} " programmer\n\ +Please describe the programmer parameters here.\n\ +\0/" \ +-e "s/.*PLACEHOLDER_NEWPROG_MAN_REQUIREMENTS/\ +.B ${PROGRAMMERNAME}\n\ +Please describe the programmer requirements here.\n\ +.sp\n\ +\0/" flashrom.8.tmpl > flashrom.8.tmpl.mine + +cat >$PROGRAMMERNAME.c.mine <<EOF +/* + * This file is part of the flashrom project. + * + * Copyright (C) $(date +%Y) ${AUTHORNAME} + * + * This program is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License as published by + * the Free Software Foundation; ${GPLV3EITHER}version 2 of the License${GPLV3OR}. + * + * This program is distributed in the hope that it will be useful, + * but WITHOUT ANY WARRANTY; without even the implied warranty of + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + * GNU General Public License for more details. + * + * You should have received a copy of the GNU General Public License + * along with this program; if not, write to the Free Software + * Foundation, Inc., 51 Franklin St, Fifth Floor, Boston, MA 02110-1301 USA + */ + +/* Driver for the ${PROGRAMMERDESCR} hardware by ${PROGRAMMERMANUF}. + * See ${PROGRAMMERURL} for more info. + */ + +#include "flash.h" +#include "programmer.h" + +EOF + +if test $DEVICETYPE = PCI -o $DEVICETYPE = USB; then +cat >>$PROGRAMMERNAME.c.mine <<EOF +const struct dev_entry devs_${PROGRAMMERNAME}[] = { + {0xdead, 0xbeef, NT, "Vendor name", "Device name"}, + + {0}, +}; + +EOF +fi + +if test $NEED_PARLPCFWH = yes; then +cat >>$PROGRAMMERNAME.c.mine <<EOF +static void ${PROGRAMMERNAME}_chip_writeb(const struct flashctx *flash, uint8_t val, chipaddr addr) +{ + /* Write a byte to the flash chip. */ +} + +static uint8_t ${PROGRAMMERNAME}_chip_readb(const struct flashctx *flash, const chipaddr addr) +{ + /* Read a byte from the flash chip and return it. */ + /* Set it to 0xff to get the template to compile. */ + uint8_t val = 0xff; + + return val; +} + +static const struct par_programmer par_programmer_${PROGRAMMERNAME} = { + .chip_readb = ${PROGRAMMERNAME}_chip_readb, + /* If your programmer supports word/long accesses, change the lines below. */ + .chip_readw = fallback_chip_readw, + .chip_readl = fallback_chip_readl, + .chip_readn = fallback_chip_readn, + .chip_writeb = ${PROGRAMMERNAME}_chip_writeb, + .chip_writew = fallback_chip_writew, + .chip_writel = fallback_chip_writel, + .chip_writen = fallback_chip_writen, +}; + +EOF +fi + +if test $NEED_SPI_BITBANG = yes; then +cat >>$PROGRAMMERNAME.c.mine <<EOF +static void ${PROGRAMMERNAME}_bitbang_set_cs(int val) +{ + /* Set/clear the CS# line. */ +} + +static void ${PROGRAMMERNAME}_bitbang_set_sck(int val) +{ + /* Set/clear the SCLK line. */ +} + +static void ${PROGRAMMERNAME}_bitbang_set_mosi(int val) +{ + /* Set/clear the MOSI line. */ +} + +static int ${PROGRAMMERNAME}_bitbang_get_miso(void) +{ + /* Get the state of the MISO line and return it. */ + /* Set it to 1 to get the template to compile. */ + int misoval = 1; + + return misoval; +} + +/* If this programmer does not support requesting/releasing the SPI bus, remove + * the functions ${PROGRAMMERNAME}_request_spibus and ${PROGRAMMERNAME}_release_spibus + * and set bitbang_spi_master_${PROGRAMMERNAME} members .request_bus and .release_bus + * to NULL. + */ +static void ${PROGRAMMERNAME}_request_spibus(void) +{ +} + +static void ${PROGRAMMERNAME}_release_spibus(void) +{ +} + +static const struct bitbang_spi_master bitbang_spi_master_${PROGRAMMERNAME} = { + .set_cs = ${PROGRAMMERNAME}_bitbang_set_cs, + .set_sck = ${PROGRAMMERNAME}_bitbang_set_sck, + .set_mosi = ${PROGRAMMERNAME}_bitbang_set_mosi, + .get_miso = ${PROGRAMMERNAME}_bitbang_get_miso, + .request_bus = ${PROGRAMMERNAME}_request_spibus, + .release_bus = ${PROGRAMMERNAME}_release_spibus, + .half_period = 1, /* Delay in microseconds before each SCLK level change. */ +}; + +EOF +fi + +if test $NEED_SPI = yes; then +cat >>$PROGRAMMERNAME.c.mine <<EOF +/* Include string.h for memset to get the template to compile. Remove this. */ +#include <string.h> +static int ${PROGRAMMERNAME}_spi_send_command(struct flashctx *flash, + unsigned int writecnt, unsigned int readcnt, + const unsigned char *writearr, + unsigned char *readarr) +{ + /* Send a SPI command to the flash chip. */ + /* Set readarr to 0xff to get the template to compile and run without segfaults. */ + memset(readarr, 0xff, readcnt); + + return 0; +} + +static const struct spi_master spi_master_${PROGRAMMERNAME} = { + .max_data_read = 64 * 1024, /* Maximum data read size in one go (excluding opcode+address). */ + .max_data_write = 256, /* Maximum data write size in one go (excluding opcode+address). */ + .command = ${PROGRAMMERNAME}_spi_send_command, + .multicommand = default_spi_send_multicommand, + .read = default_spi_read, + .write_256 = default_spi_write_256, + .write_aai = default_spi_write_aai, +}; + +EOF +fi + +cat >>$PROGRAMMERNAME.c.mine <<EOF +static int ${PROGRAMMERNAME}_shutdown(void *data) +{ + /* Shutdown stuff. */ + return 0; +} + +int ${PROGRAMMERNAME}_init(void) +{ + /* Init stuff (i.e. parameter parsing) here which does not need to be + * undone. + */ + + /* If your shutdown function takes a parameter, replace NULL with it. */ + register_shutdown(${PROGRAMMERNAME}_shutdown, NULL); + + /* Init stuff which needs to be undone on shutdown. */ + +EOF + +if test $NEED_SPI_BITBANG = yes; then +cat >>$PROGRAMMERNAME.c.mine <<EOF + /* 1 usec halfperiod delay, change as needed. */ + if (bitbang_spi_init(&bitbang_spi_master_${PROGRAMMERNAME})) + return 1; + +EOF +fi + +if test $NEED_SPI = yes; then +cat >>$PROGRAMMERNAME.c.mine <<EOF + register_spi_master(&spi_master_${PROGRAMMERNAME}); + +EOF +fi + +if test $NEED_PARLPCFWH = yes; then +cat >>$PROGRAMMERNAME.c.mine <<EOF + register_par_programmer(&par_programmer_${PROGRAMMERNAME}, BUS_${BUS_PARLPCFWH}); + +EOF +fi + +cat >>$PROGRAMMERNAME.c.mine <<EOF + return 0; +} +EOF + +csplit -f .newmeson_options meson_options.txt "/#PLACEHOLDER_NEWPROG_MESON_OPTION_START/+1" "/#PLACEHOLDER_NEWPROG_MESON_OPTION_END/" +echo "option('config_${PROGRAMMERNAME}', type : 'boolean', value : true, description : '${PROGRAMMERDESCR}')" >>.newmeson_options01 +sort .newmeson_options01 >.newmeson_options03 +cat .newmeson_options00 .newmeson_options03 .newmeson_options02 >meson_options.txt.mine +rm .newmeson_options00 .newmeson_options01 .newmeson_options02 .newmeson_options03 + +csplit -f .newmeson meson.build "/#PLACEHOLDER_NEWPROG_MESON_CONFIGFETCH_START/+1" "/#PLACEHOLDER_NEWPROG_MESON_CONFIGFETCH_END/" +echo "config_${PROGRAMMERNAME} = get_option('config_${PROGRAMMERNAME}')" >>.newmeson01 +sort .newmeson01 >.newmeson03 +cat .newmeson00 .newmeson03 .newmeson02 >.newmeson.build.mine +rm .newmeson00 .newmeson01 .newmeson02 .newmeson03 + +if test $DEVICETYPE = PCI ; then + csplit -f .newmeson .newmeson.build.mine "/#PLACEHOLDER_NEWPROG_MESON_PCI_REQUIREMENT_MISSING_START/+1" "/#PLACEHOLDER_NEWPROG_MESON_PCI_REQUIREMENT_MISSING_END/" + echo " config_${PROGRAMMERNAME} = false" >>.newmeson01 + sort .newmeson01 >.newmeson03 + cat .newmeson00 .newmeson03 .newmeson02 >.newmeson.build.mine + rm .newmeson00 .newmeson01 .newmeson02 .newmeson03 +fi + +if test $DEVICETYPE = USB ; then + csplit -f .newmeson .newmeson.build.mine "/#PLACEHOLDER_NEWPROG_MESON_USB_REQUIREMENT_MISSING_START/+1" "/#PLACEHOLDER_NEWPROG_MESON_USB_REQUIREMENT_MISSING_END/" + echo " config_${PROGRAMMERNAME} = false" >>.newmeson01 + sort .newmeson01 >.newmeson03 + cat .newmeson00 .newmeson03 .newmeson02 >.newmeson.build.mine + rm .newmeson00 .newmeson01 .newmeson02 .newmeson03 +fi + +csplit -f .newmeson .newmeson.build.mine "/#PLACEHOLDER_NEWPROG_MESON_FILES_DEFINES_NEEDS_START/+1" "/#PLACEHOLDER_NEWPROG_MESON_FILES_DEFINES_NEEDS_END/" +# FIXME: The current meson.build always builds the PCI intrastructure unless explicitly disabled. +cat >>.newmeson01 <<EOF +if config_${PROGRAMMERNAME} + srcs += '${PROGRAMMERNAME}.c' + cargs += '-D${CONFIGNAME}=1' +EOF +if $NEED_SERIAL = yes; then + cat >>.newmeson01 <<EOF + need_serial = true +EOF +cat >>.newmeson01 <<EOF +endif +EOF +# FIXME: Sorting is a bit more complicated here. Skip it for now. +cat .newmeson00 .newmeson01 .newmeson02 >.newmeson.build.mine +rm .newmeson00 .newmeson01 .newmeson02 + +mv .newmeson.build.mine meson.build.mine + +echo "The driver skeleton has been created in $PROGRAMMERNAME.c.mine" +echo "Modified versions of existing files have been created with extension .mine" +echo "You can replace the original files with the modified versions by running" +echo "for a in *; do test -f $a.mine && mv $a.mine $a; done" +echo "If you want to use the newly generated skeleton $PROGRAMMERNAME.c.mine , run" +echo "mv $PROGRAMMERNAME.c.mine $PROGRAMMERNAME.c" +echo +echo "WARNING: Please note that rerunning build_new_driver.sh will overwrite" +echo "all *.mine files, but it won't touch $PROGRAMMERNAME.c ." +echo "If something goes wrong, you can revert all files which look odd and" +echo "run this script again." diff --git a/flashrom.8.tmpl b/flashrom.8.tmpl index aa5bcd4..b28c86a 100644 --- a/flashrom.8.tmpl +++ b/flashrom.8.tmpl @@ -345,6 +345,7 @@ .sp .BR "* stlinkv3_spi" " (for SPI flash ROMs attached to STMicroelectronics STLINK V3 devices)" .sp +."PLACEHOLDER_NEWPROG_MAN_SHORTDESCRIPTION Some programmers have optional or mandatory parameters which are described in detail in the .B PROGRAMMER-SPECIFIC INFORMATION @@ -1287,7 +1288,7 @@ If the passed frequency is not supported by the adapter the nearest lower supported frequency will be used. .SS - +."PLACEHOLDER_NEWPROG_MAN_LONGDESCRIPTION .SH EXAMPLES To back up and update your BIOS, run .sp @@ -1365,6 +1366,7 @@ .B ogp needs PCI configuration space read access and raw memory access. .sp +."PLACEHOLDER_NEWPROG_MAN_REQUIREMENTS On OpenBSD, you can obtain raw access permission by setting .B "securelevel=-1" in diff --git a/flashrom.c b/flashrom.c index e540027..ad8b23e 100644 --- a/flashrom.c +++ b/flashrom.c @@ -473,6 +473,7 @@ }, #endif
+//PLACEHOLDER_NEWPROG_PROGRAMMER_ARRAY {0}, /* This entry corresponds to PROGRAMMER_INVALID. */ };
diff --git a/meson.build b/meson.build index 375089c..c602399 100644 --- a/meson.build +++ b/meson.build @@ -30,6 +30,7 @@ add_project_arguments('-DFLASHROM_VERSION="' + meson.project_version() + '"', language : 'c')
# get defaults from configure +# PLACEHOLDER_NEWPROG_MESON_CONFIGFETCH_START config_atahpt = get_option('config_atahpt') config_atapromise = get_option('config_atapromise') config_atavia = get_option('config_atavia') @@ -62,6 +63,7 @@ config_serprog = get_option('config_serprog') config_usbblaster_spi = get_option('config_usbblaster_spi') config_stlinkv3_spi = get_option('config_stlinkv3_spi') +# PLACEHOLDER_NEWPROG_MESON_CONFIGFETCH_END
cargs = [] deps = [] @@ -86,11 +88,13 @@ srcs += 'usbdev.c' deps += dependency('libusb-1.0') else +# PLACEHOLDER_NEWPROG_MESON_USB_REQUIREMENT_MISSING_START config_ch341a_spi = false config_dediprog = false config_digilent_spi = false config_developerbox_spi = false config_pickit2_spi = false +# PLACEHOLDER_NEWPROG_MESON_USB_REQUIREMENT_MISSING_END endif
# some programmers require libpci @@ -99,6 +103,7 @@ deps += dependency('libpci') cargs += '-DNEED_PCI=1' else +# PLACEHOLDER_NEWPROG_MESON_PCI_REQUIREMENT_MISSING_START config_atahpt = false config_atapromise = false config_atavia = false @@ -116,9 +121,11 @@ config_rayer_spi = false config_satamv = false config_satasii = false +# PLACEHOLDER_NEWPROG_MESON_PCI_REQUIREMENT_MISSING_END endif
# set defines for configured programmers +# PLACEHOLDER_NEWPROG_MESON_FILES_DEFINES_NEEDS_START if config_atahpt srcs += 'atahpt.c' cargs += '-DCONFIG_ATAHPT=1' @@ -276,6 +283,7 @@ srcs += 'stlinkv3_spi.c' cargs += '-DCONFIG_STLINKV3_SPI=1' endif +# PLACEHOLDER_NEWPROG_MESON_FILES_DEFINES_NEEDS_END
# bitbanging SPI infrastructure if config_bitbang_spi diff --git a/meson_options.txt b/meson_options.txt index b7633d0..9e47272 100644 --- a/meson_options.txt +++ b/meson_options.txt @@ -1,6 +1,7 @@ option('pciutils', type : 'boolean', value : true, description : 'use pciutils') option('usb', type : 'boolean', value : true, description : 'use libusb1')
+#PLACEHOLDER_NEWPROG_MESON_OPTION_START option('config_atahpt', type : 'boolean', value : false, description : 'Highpoint (HPT) ATA/RAID controllers') option('config_atapromise', type : 'boolean', value : false, description : 'Promise ATA controller') option('config_atavia', type : 'boolean', value : true, description : 'VIA VT6421A LPC memory') @@ -33,3 +34,4 @@ option('config_satasii', type : 'boolean', value : true, description : 'SiI SATA controllers') option('config_serprog', type : 'boolean', value : true, description : 'serprog') option('config_usbblaster_spi', type : 'boolean', value : true, description : 'Altera USB-Blaster dongles') +#PLACEHOLDER_NEWPROG_MESON_OPTION_END diff --git a/programmer.h b/programmer.h index 3cf53b9..90afc7b 100644 --- a/programmer.h +++ b/programmer.h @@ -127,6 +127,7 @@ #if CONFIG_STLINKV3_SPI == 1 PROGRAMMER_STLINKV3_SPI, #endif +//PLACEHOLDER_NEWPROG_PROGRAMMER_ENUM PROGRAMMER_INVALID /* This must always be the last entry. */ };
@@ -573,6 +574,8 @@ int ni845x_spi_init(void); #endif
+//PLACEHOLDER_NEWPROG_PUBLICFUNCTIONS + /* flashrom.c */ struct decode_sizes { uint32_t parallel;
Angel Pons has posted comments on this change. ( https://review.coreboot.org/c/flashrom/+/38488 )
Change subject: Automatic programmer driver writer ......................................................................
Patch Set 1:
(1 comment)
That sounds pretty interesting
https://review.coreboot.org/c/flashrom/+/38488/1//COMMIT_MSG Commit Message:
https://review.coreboot.org/c/flashrom/+/38488/1//COMMIT_MSG@28 PS1, Line 28: Icdcdcdcdcdcdcdcdcdcdcdcdcdcdcdcdcd000001 That is a weird Change-Id :D
Idwer Vollering has posted comments on this change. ( https://review.coreboot.org/c/flashrom/+/38488 )
Change subject: Automatic programmer driver writer ......................................................................
Patch Set 1:
(2 comments)
https://review.coreboot.org/c/flashrom/+/38488/1/build_new_driver.sh File build_new_driver.sh:
https://review.coreboot.org/c/flashrom/+/38488/1/build_new_driver.sh@1 PS1, Line 1: #!/bin/bash '/usr/bin/env bash' is preferred.
https://review.coreboot.org/c/flashrom/+/38488/1/build_new_driver.sh@42 PS1, Line 42: What are your thoughts on putting the lines below in a separate file (build_new_driver.sh.inc)?
Paul Menzel has posted comments on this change. ( https://review.coreboot.org/c/flashrom/+/38488 )
Change subject: Automatic programmer driver writer ......................................................................
Patch Set 1:
(1 comment)
https://review.coreboot.org/c/flashrom/+/38488/1//COMMIT_MSG Commit Message:
https://review.coreboot.org/c/flashrom/+/38488/1//COMMIT_MSG@16 PS1, Line 16: for a in *; do test -f $a.mine && mv $a.mine $a; done Shouldn’t we rely on the VCS to list the changes, and list newly added files (`git status`)?
Hello build bot (Jenkins),
I'd like you to reexamine a change. Please visit
https://review.coreboot.org/c/flashrom/+/38488
to look at the new patch set (#2).
Change subject: Automatic programmer driver writer ......................................................................
Automatic programmer driver writer
Copy ezo-driver.tmpl to a file of your choice and fill in the properties of the flashrom programmer driver you want to create. Then run ./build_new_driver.sh <your-driver.tmpl>
The driver skeleton will be created in $PROGRAMMERNAME.c.mine Modified versions of existing files will be created with extension .mine You can replace the original files with the modified versions by running for a in *; do test -f $a.mine && mv $a.mine $a; done
If you want to use the newly generated skeleton $PROGRAMMERNAME.c.mine , run mv $PROGRAMMERNAME.c.mine $PROGRAMMERNAME.c
WARNING: Please note that rerunning build_new_driver.sh will overwrite all *.mine files, but it won't touch $PROGRAMMERNAME.c . If something goes wrong, you can revert all files which look odd and run this script again.
Change-Id: Icdcdcdcdcdcdcdcdcdcdcdcdcdcdcdcdcd000001 Signed-off-by: Carl-Daniel Hailfinger c-d.hailfinger.devel.2006@gmx.net --- M Makefile A build_new_driver.sh A ezo-driver.tmpl M flashrom.8.tmpl M flashrom.c M meson.build M meson_options.txt M programmer.h 8 files changed, 475 insertions(+), 1 deletion(-)
git pull ssh://review.coreboot.org:29418/flashrom refs/changes/88/38488/2
Carl-Daniel Hailfinger has posted comments on this change. ( https://review.coreboot.org/c/flashrom/+/38488 )
Change subject: Automatic programmer driver writer ......................................................................
Patch Set 2:
(2 comments)
https://review.coreboot.org/c/flashrom/+/38488/1//COMMIT_MSG Commit Message:
https://review.coreboot.org/c/flashrom/+/38488/1//COMMIT_MSG@16 PS1, Line 16: for a in *; do test -f $a.mine && mv $a.mine $a; done
Shouldn’t we rely on the VCS to list the changes, and list newly added files (`git status`)?
No, that would also mess with files which you already changed manually. And a revert would then kill both the manual changes as well as the automatic changes. And running git status is not fun, especially when you're trying to write a driver and have other data files in the same directory.
https://review.coreboot.org/c/flashrom/+/38488/1//COMMIT_MSG@28 PS1, Line 28: Icdcdcdcdcdcdcdcdcdcdcdcdcdcdcdcdcd000001
That is a weird Change-Id :D
Indeed, the Change-Id is freely selectable subject to some length and character set constraints. I might want to change it later.
Carl-Daniel Hailfinger has posted comments on this change. ( https://review.coreboot.org/c/flashrom/+/38488 )
Change subject: Automatic programmer driver writer ......................................................................
Patch Set 2:
(2 comments)
Thanks for the comments.
https://review.coreboot.org/c/flashrom/+/38488/1/build_new_driver.sh File build_new_driver.sh:
https://review.coreboot.org/c/flashrom/+/38488/1/build_new_driver.sh@1 PS1, Line 1: #!/bin/bash
'/usr/bin/env bash' is preferred.
Will be fixed in the next iteration.
https://review.coreboot.org/c/flashrom/+/38488/1/build_new_driver.sh@42 PS1, Line 42:
What are your thoughts on putting the lines below in a separate file (build_new_driver.sh. […]
Done.
Hello build bot (Jenkins),
I'd like you to reexamine a change. Please visit
https://review.coreboot.org/c/flashrom/+/38488
to look at the new patch set (#3).
Change subject: Automatic programmer driver writer ......................................................................
Automatic programmer driver writer
Copy ezo-driver.tmpl to a file of your choice and fill in the properties of the flashrom programmer driver you want to create. Then run ./build_new_driver.sh <your-driver.tmpl>
The driver skeleton will be created in $PROGRAMMERNAME.c.mine Modified versions of existing files will be created with extension .mine You can replace the original files with the modified versions by running for a in *; do test -f $a.mine && mv $a.mine $a; done
If you want to use the newly generated skeleton $PROGRAMMERNAME.c.mine , run mv $PROGRAMMERNAME.c.mine $PROGRAMMERNAME.c
WARNING: Please note that rerunning build_new_driver.sh will overwrite all *.mine files, but it won't touch $PROGRAMMERNAME.c . If something goes wrong, you can revert all files which look odd and run this script again.
Change-Id: Icdcdcdcdcdcdcdcdcdcdcdcdcdcdcdcdcd000001 Signed-off-by: Carl-Daniel Hailfinger c-d.hailfinger.devel.2006@gmx.net --- M Makefile A build_new_driver.sh A ezo-driver.tmpl M flashrom.8.tmpl M flashrom.c M meson.build M meson_options.txt M programmer.h 8 files changed, 475 insertions(+), 1 deletion(-)
git pull ssh://review.coreboot.org:29418/flashrom refs/changes/88/38488/3
Carl-Daniel Hailfinger has posted comments on this change. ( https://review.coreboot.org/c/flashrom/+/38488 )
Change subject: Automatic programmer driver writer ......................................................................
Patch Set 3:
(4 comments)
AFAICS all comments have now been addressed.
https://review.coreboot.org/c/flashrom/+/38488/1//COMMIT_MSG Commit Message:
https://review.coreboot.org/c/flashrom/+/38488/1//COMMIT_MSG@16 PS1, Line 16: for a in *; do test -f $a.mine && mv $a.mine $a; done
No, that would also mess with files which you already changed manually. […]
Done
https://review.coreboot.org/c/flashrom/+/38488/1//COMMIT_MSG@28 PS1, Line 28: Icdcdcdcdcdcdcdcdcdcdcdcdcdcdcdcdcd000001
Indeed, the Change-Id is freely selectable subject to some length and character set constraints. […]
Done
https://review.coreboot.org/c/flashrom/+/38488/1/build_new_driver.sh File build_new_driver.sh:
https://review.coreboot.org/c/flashrom/+/38488/1/build_new_driver.sh@1 PS1, Line 1: #!/bin/bash
Will be fixed in the next iteration.
Done
https://review.coreboot.org/c/flashrom/+/38488/1/build_new_driver.sh@42 PS1, Line 42:
Done.
Done
Paul Menzel has posted comments on this change. ( https://review.coreboot.org/c/flashrom/+/38488 )
Change subject: Automatic programmer driver writer ......................................................................
Patch Set 3:
(2 comments)
https://review.coreboot.org/c/flashrom/+/38488/1//COMMIT_MSG Commit Message:
https://review.coreboot.org/c/flashrom/+/38488/1//COMMIT_MSG@16 PS1, Line 16: for a in *; do test -f $a.mine && mv $a.mine $a; done
Done
Then you are using git incorrectly.
https://review.coreboot.org/c/flashrom/+/38488/1//COMMIT_MSG@28 PS1, Line 28: Icdcdcdcdcdcdcdcdcdcdcdcdcdcdcdcdcd000001
Done
You shouldn’t change it later to not confuse Gerrit.
Carl-Daniel Hailfinger has posted comments on this change. ( https://review.coreboot.org/c/flashrom/+/38488 )
Change subject: Automatic programmer driver writer ......................................................................
Patch Set 3:
(2 comments)
https://review.coreboot.org/c/flashrom/+/38488/1//COMMIT_MSG Commit Message:
https://review.coreboot.org/c/flashrom/+/38488/1//COMMIT_MSG@16 PS1, Line 16: for a in *; do test -f $a.mine && mv $a.mine $a; done
Then you are using git incorrectly.
I would love to use git correctly. How can I make git aware that I do not want to add the same chunk twice to a file when I rerun build_new_driver.sh ? Right now you can run the script as often as you want without having the new chunks inserted multiple times.
https://review.coreboot.org/c/flashrom/+/38488/1//COMMIT_MSG@28 PS1, Line 28: Icdcdcdcdcdcdcdcdcdcdcdcdcdcdcdcdcd000001
You shouldn’t change it later to not confuse Gerrit.
Oh, the Change-Id gets changed all the time, just not explicitly. People abandon their change and submit a new one. The effect on Gerrit is identical (history is lost). I do agree that losing review history is suboptimal, though. I might skip it this time.
Hello build bot (Jenkins),
I'd like you to reexamine a change. Please visit
https://review.coreboot.org/c/flashrom/+/38488
to look at the new patch set (#4).
Change subject: Automatic programmer driver writer ......................................................................
Automatic programmer driver writer
Copy ezo-driver.tmpl to a file of your choice and fill in the properties of the flashrom programmer driver you want to create. Then run ./build_new_driver.sh <your-driver.tmpl>
The driver skeleton will be created in $PROGRAMMERNAME.c.mine Modified versions of existing files will be created with extension .mine You can replace the original files with the modified versions by running for a in *; do test -f $a.mine && mv $a.mine $a; done
FIXME: This kills the driver.c if it already exists.
If you want to use the newly generated skeleton $PROGRAMMERNAME.c.mine , run mv $PROGRAMMERNAME.c.mine $PROGRAMMERNAME.c
WARNING: Please note that rerunning build_new_driver.sh will overwrite all *.mine files, but it won't touch $PROGRAMMERNAME.c . If something goes wrong, you can revert all files which look odd and run this script again.
Also included is a Python 3 script reorder-meson.py which checks the meson build files for consistency and alphabetical ordering. Fix the alphabetical ordering of the meson files while we're at it.
Change-Id: Icdcdcdcdcdcdcdcdcdcdcdcdcdcdcdcdcd000001 Signed-off-by: Carl-Daniel Hailfinger c-d.hailfinger.devel.2006@gmx.net --- M Makefile A build_new_driver.sh A ezo-driver.tmpl M flashrom.8.tmpl M flashrom.c M meson.build M meson_options.txt M programmer.h A reorder-meson.py 9 files changed, 599 insertions(+), 15 deletions(-)
git pull ssh://review.coreboot.org:29418/flashrom refs/changes/88/38488/4
Carl-Daniel Hailfinger has posted comments on this change. ( https://review.coreboot.org/c/flashrom/+/38488 )
Change subject: Automatic programmer driver writer ......................................................................
Patch Set 4:
This changeset has two functions: build_new_driver.sh creates a skeleton for a new driver and hooks it up in all the right right places. reorder-meson.py cross-checks the meson build files against each other and fixes alphabetical ordering in both. In theory, this script could also be used as a commit hook with some minor changes.
-
Angel Pons (Code Review) -
Carl-Daniel Hailfinger (Code Review)
-
Idwer Vollering (Code Review)
-
Paul Menzel (Code Review)