Use logging feature instead of FPGA_DEBUG
v4 cover Use logging feature instead of FPGA_DEBUG
Hei hei,
while working on FPGA support for a new device I discovered debug logging in some FPGA drivers is still done as in the old days. Bring that to what I thougt would be the currently preferred approach.
Notes: Adding those Kconfig symbols in patch 1 is just to be able to build two of the old drivers.
All drivers touched were build tested with sandbox_defconfig and GCC 8 on Debian GNU/Linux 10 (buster).
Lines with other possibly questionable output were not touched, only what seemed to be designated debug output, and only for FPGA drivers having that ancient FPGA_DEBUG / PRINTF macros, so there's room for future improvements.
Changelog:
v3 -> v4:
- Reordered patches, Kconfig patch comes first now (made it easier to build and test the series step by step)
- Added three patches fixing printf compiler warnings first, before changing to the new logging framework (so CI should not fail anymore with -Werror)
v2 -> v3:
- Patch introducing FPGA uclass was completely reworked, sent independently from this series, and applied already, thus removed
- Because requiring that new FPGA uclass changes, rebased on Michal's microblaze branch '20221005'
- Removed '"%s …", __func__' and '"%d …", __line__' from log messages, because log framework can add those (enabled by CONFIG_LOGF_FUNC and CONFIG_LOGF_LINE)
v1 -> v2:
- Rebased on master
- Added patch to introduce new FPGA uclass in front of the other patches
- Use that new uclass as log category
- Slightly reworded cover letter
Greets
Alex
Alexander Dahl (10):
fpga: Add missing Kconfig symbols for old FPGA drivers
fpga: spartan2: Fix printf arguments warning
fpga: spartan3: Fix printf arguments warning
fpga: virtex2: Fix printf format string warnings
fpga: altera: Use logging feature instead of FPGA_DEBUG
fpga: cyclon2: Use logging feature instead of FPGA_DEBUG
fpga: ACEX1K: Use logging feature instead of FPGA_DEBUG
fpga: spartan2: Use logging feature instead of FPGA_DEBUG
fpga: spartan3: Use logging feature instead of FPGA_DEBUG
fpga: virtex2: Use logging feature instead of FPGA_DEBUG
drivers/fpga/ACEX1K.c | 37 +++++++++---------- drivers/fpga/Kconfig | 12 +++++++ drivers/fpga/altera.c | 11 +++--- drivers/fpga/cyclon2.c | 38 +++++++++----------- drivers/fpga/spartan2.c | 80 +++++++++++++++++++---------------------- drivers/fpga/spartan3.c | 80 +++++++++++++++++++---------------------- drivers/fpga/virtex2.c | 69 ++++++++++++++++------------------- 7 files changed, 152 insertions(+), 175 deletions(-)
base-commit: 2d8cf392a77815f062446ef441f1078958dc1b2a
Signature · 1 line
2.30.2
PATCH v4 1/10 fpga: Add missing Kconfig symbols for old FPGA drivers
Those drivers could not be built anymore without those options present.
--- drivers/fpga/Kconfig | 12 ++++++++++++ 1 file changed, 12 insertions(+)
drivers/fpga/Kconfig +12 -0
@@ -27,6 +27,12 @@ config FPGA_STRATIX_V help Say Y here to enable the Altera Stratix V FPGA specific driver. +config FPGA_ACEX1K + bool "Enable Altera ACEX 1K driver" + depends on FPGA_ALTERA + help + Say Y here to enable the Altera ACEX 1K FPGA specific driver. + config FPGA_CYCLON2 bool "Enable Altera FPGA driver for Cyclone II" depends on FPGA_ALTERA
@@ -71,6 +77,12 @@ config FPGA_VERSALPL Versal. The bitstream will only be generated as PDI for Versal platform. +config FPGA_SPARTAN2 + bool "Enable Spartan2 FPGA driver" + depends on FPGA_XILINX + help + Enable Spartan2 FPGA driver. + config FPGA_SPARTAN3 bool "Enable Spartan3 FPGA driver" depends on FPGA_XILINX
Signature · 1 line
2.30.2
PATCH v4 2/10 fpga: spartan2: Fix printf arguments warning
In reply to Alexander DahlThat extra comma messes up format arguments.
Warning appears if built with FPGA_DEBUG defined:
CC drivers/fpga/spartan2.o
/mnt/data/adahl/src/u-boot/drivers/fpga/spartan2.c: In function ‘spartan2_sp_load’:
/mnt/data/adahl/src/u-boot/drivers/fpga/spartan2.c:112:11: warning: too many arguments for format [-Wformat-extra-args]
PRINTF ("%s: Function Table:\n"
^~~~~~~~~~~~~~~~~~~~~~~
/mnt/data/adahl/src/u-boot/drivers/fpga/spartan2.c:12:37: note: in definition of macro ‘PRINTF’
#define PRINTF(fmt,args...) printf (fmt ,##args)
^~~
CC drivers/fpga/spartan3.o
/mnt/data/adahl/src/u-boot/drivers/fpga/spartan3.c: In function ‘spartan3_sp_load’:
/mnt/data/adahl/src/u-boot/drivers/fpga/spartan3.c:117:11: warning: too many arguments for format [-Wformat-extra-args]
PRINTF ("%s: Function Table:\n"
^~~~~~~~~~~~~~~~~~~~~~~
/mnt/data/adahl/src/u-boot/drivers/fpga/spartan3.c:17:37: note: in definition of macro ‘PRINTF’
#define PRINTF(fmt,args...) printf (fmt ,##args)
^~~
--- drivers/fpga/spartan2.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-)
drivers/fpga/spartan2.c +1 -1
@@ -122,7 +122,7 @@ static int spartan2_sp_load(xilinx_desc *desc, const void *buf, size_t bsize)"read data:\t0x%p\n" "write data:\t0x%p\n" "busy:\t0x%p\n" -"abort:\t0x%p\n" , +"abort:\t0x%p\n" "post:\t0x%p\n\n" , __FUNCTION__, &fn, fn, fn->pre, fn->pgm, fn->init, fn->err, fn->clk, fn->cs, fn->wr, fn->rdata, fn->wdata, fn->busy,
Signature · 1 line
2.30.2
PATCH v4 3/10 fpga: spartan3: Fix printf arguments warning
In reply to Alexander DahlThe additional comma messes up the arguments.
Warning appears if built with FPGA_DEBUG defined:
CC drivers/fpga/spartan3.o
/mnt/data/adahl/src/u-boot/drivers/fpga/spartan3.c: In function ‘spartan3_sp_load’:
/mnt/data/adahl/src/u-boot/drivers/fpga/spartan3.c:118:11: warning: too many arguments for format [-Wformat-extra-args]
PRINTF ("%s: Function Table:\n"
^~~~~~~~~~~~~~~~~~~~~~~
/mnt/data/adahl/src/u-boot/drivers/fpga/spartan3.c:18:37: note: in definition of macro ‘PRINTF’
#define PRINTF(fmt,args...) printf (fmt ,##args)
^~~
--- drivers/fpga/spartan3.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-)
drivers/fpga/spartan3.c +1 -1
@@ -127,7 +127,7 @@ static int spartan3_sp_load(xilinx_desc *desc, const void *buf, size_t bsize)"read data:\t0x%p\n" "write data:\t0x%p\n" "busy:\t0x%p\n" -"abort:\t0x%p\n" , +"abort:\t0x%p\n" "post:\t0x%p\n\n" , __FUNCTION__, &fn, fn, fn->pre, fn->pgm, fn->init, fn->err, fn->clk, fn->cs, fn->wr, fn->rdata, fn->wdata, fn->busy,
Signature · 1 line
2.30.2
PATCH v4 4/10 fpga: virtex2: Fix printf format string warnings
In reply to Alexander DahlWarning appears if built with FPGA_DEBUG defined:
CC drivers/fpga/virtex2.o/mnt/data/adahl/src/u-boot/drivers/fpga/virtex2.c: In function ‘virtex2_ssm_load’:/mnt/data/adahl/src/u-boot/drivers/fpga/virtex2.c:333:11: warning: format ‘%d’ expects argument of type ‘int’, but argument 4 has type ‘size_t’ {aka ‘long unsigned int’} [-Wformat=]
PRINTF("%s:%d:done went active early, bytecount = %d\n",
^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
__func__, __LINE__, bytecount);
~~~~~~~~~
/mnt/data/adahl/src/u-boot/drivers/fpga/virtex2.c:25:37: note: in definition of macro ‘PRINTF’
#define PRINTF(fmt, args...) printf(fmt, ##args)
^~~
/mnt/data/adahl/src/u-boot/drivers/fpga/virtex2.c: In function ‘virtex2_ss_load’:
/mnt/data/adahl/src/u-boot/drivers/fpga/virtex2.c:468:12: warning: format ‘%d’ expects argument of type ‘int’, but argument 4 has type ‘size_t’ {aka ‘long unsigned int’} [-Wformat=]
PRINTF("%s:%d:done went active early, bytecount = %d\n",
^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
__func__, __LINE__, bytecount);
~~~~~~~~~
/mnt/data/adahl/src/u-boot/drivers/fpga/virtex2.c:25:37: note: in definition of macro ‘PRINTF’
#define PRINTF(fmt, args...) printf(fmt, ##args)
^~~--- drivers/fpga/virtex2.c | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-)
drivers/fpga/virtex2.c +2 -2
@@ -330,7 +330,7 @@ static int virtex2_ssm_load(xilinx_desc *desc, const void *buf, size_t bsize)#endif if ((*fn->done)(cookie) ==FPGA_SUCCESS ) { -PRINTF ("%s:%d:done went active early, bytecount = %d\n" , +PRINTF ("%s:%d:done went active early, bytecount = %zu\n" , __func__, __LINE__, bytecount);break ; }
@@ -465,7 +465,7 @@ static int virtex2_ss_load(xilinx_desc *desc, const void *buf, size_t bsize)#endif if ((*fn->done)(cookie) ==FPGA_SUCCESS ) { -PRINTF ("%s:%d:done went active early, bytecount = %d\n" , +PRINTF ("%s:%d:done went active early, bytecount = %zu\n" , __func__, __LINE__, bytecount);break ; }
Signature · 1 line
2.30.2
PATCH v4 5/10 fpga: altera: Use logging feature instead of FPGA_DEBUG
In reply to Alexander DahlInstead of using DEBUG or LOG_DEBUG the driver still had its own definition for debug output.
--- drivers/fpga/altera.c | 11 ++++------- 1 file changed, 4 insertions(+), 7 deletions(-)
drivers/fpga/altera.c +4 -7
@@ -7,6 +7,8 @@* Rich Ireland, Enterasys Networks, [email protected]. */ +#define LOG_CATEGORY UCLASS_FPGA +/* * Altera FPGA support */
@@ -16,9 +18,6 @@#include <log.h> #include <stratixII.h> -/* Define FPGA_DEBUG to 1 to get debug printf's */ -#define FPGA_DEBUG 0 - static const struct altera_fpga { enum altera_family family; const char *name;
@@ -106,8 +105,7 @@ int altera_load(Altera_desc *desc, const void *buf, size_t bsize)if (!fpga)return FPGA_FAIL ; -debug_cond (FPGA_DEBUG ,"%s: Launching the %s Loader...\n" , - __func__, fpga->name); +log_debug ("Launching the %s Loader...\n" , fpga->name);if (fpga->load)return fpga->load (desc, buf, bsize);return 0 ;
@@ -120,8 +118,7 @@ int altera_dump(Altera_desc *desc, const void *buf, size_t bsize)if (!fpga)return FPGA_FAIL ; -debug_cond (FPGA_DEBUG ,"%s: Launching the %s Reader...\n" , - __func__, fpga->name); +log_debug ("Launching the %s Reader...\n" , fpga->name);if (fpga->dump)return fpga->dump (desc, buf, bsize);return 0 ;
Signature · 1 line
2.30.2
PATCH v4 6/10 fpga: cyclon2: Use logging feature instead of FPGA_DEBUG
In reply to Alexander DahlInstead of using DEBUG or LOG_DEBUG the driver still had its own definition for debug output.
--- drivers/fpga/cyclon2.c | 38 ++++++++++++++++---------------------- 1 file changed, 16 insertions(+), 22 deletions(-)
drivers/fpga/cyclon2.c +16 -22
@@ -5,18 +5,14 @@* Based on ACE1XK.c */ +#define LOG_CATEGORY UCLASS_FPGA +#include <common.h> /* core U-Boot definitions */ +#include <log.h> #include <altera.h> #include <ACEX1K.h> /* ACEX device family */ #include <linux/delay.h> -/* Define FPGA_DEBUG to get debug printf's */ -#ifdef FPGA_DEBUG -#define PRINTF(fmt, args...) printf(fmt, ##args) -#else -#define PRINTF(fmt, args...) -#endif -/* Note: The assumption is that we cannot possibly run fast enough to * overrun the device (the Slave Parallel mode can free run at 50MHz). * If there is a need to operate slower, define CONFIG_FPGA_DELAY in
@@ -42,7 +38,7 @@ int CYC2_load(Altera_desc *desc, const void *buf, size_t bsize)switch (desc->iface) {case passive_serial: -PRINTF ("%s: Launching Passive Serial Loader\n" , __func__); +log_debug ("Launching Passive Serial Loader\n" ); ret_val =CYC2_ps_load (desc, buf, bsize);break ;
@@ -51,8 +47,7 @@ int CYC2_load(Altera_desc *desc, const void *buf, size_t bsize)* done in the write() callback. Use the existing PS load * function for FPP, too. */ -PRINTF ("%s: Launching Fast Passive Parallel Loader\n" , - __func__); +log_debug ("Launching Fast Passive Parallel Loader\n" ); ret_val =CYC2_ps_load (desc, buf, bsize);break ;
@@ -72,7 +67,7 @@ int CYC2_dump(Altera_desc *desc, const void *buf, size_t bsize)switch (desc->iface) {case passive_serial: -PRINTF ("%s: Launching Passive Serial Dump\n" , __func__); +log_debug ("Launching Passive Serial Dump\n" ); ret_val =CYC2_ps_dump (desc, buf, bsize);break ;
@@ -99,22 +94,21 @@ static int CYC2_ps_load(Altera_desc *desc, const void *buf, size_t bsize)Altera_CYC2_Passive_Serial_fns *fn = desc->iface_fns;int ret =0 ; -PRINTF ("%s: start with interface functions @ 0x%p\n" , - __func__, fn); +log_debug ("start with interface functions @ 0x%p\n" , fn);if (fn) {int cookie = desc->cookie;/* make a local copy */ unsigned long ts;/* timestamp */ -PRINTF ("%s: Function Table:\n" -"ptr:\t0x%p\n" -"struct: 0x%p\n" -"config:\t0x%p\n" -"status:\t0x%p\n" -"write:\t0x%p\n" -"done:\t0x%p\n\n" , - __func__, &fn, fn, fn->config, fn->status, - fn->write, fn->done); +log_debug ("Function Table:\n" +"ptr:\t0x%p\n" +"struct: 0x%p\n" +"config:\t0x%p\n" +"status:\t0x%p\n" +"write:\t0x%p\n" +"done:\t0x%p\n\n" , + &fn, fn, fn->config, fn->status, + fn->write, fn->done);#ifdef CONFIG_SYS_FPGA_PROG_FEEDBACK printf ("Loading FPGA Device %d..." , cookie);#endif
Signature · 1 line
2.30.2
PATCH v4 7/10 fpga: ACEX1K: Use logging feature instead of FPGA_DEBUG
In reply to Alexander DahlInstead of using DEBUG or LOG_DEBUG the driver still had its own definition for debug output.
--- drivers/fpga/ACEX1K.c | 37 ++++++++++++++++--------------------- 1 file changed, 16 insertions(+), 21 deletions(-)
drivers/fpga/ACEX1K.c +16 -21
@@ -7,18 +7,14 @@* Rich Ireland, Enterasys Networks, [email protected]. */ +#define LOG_CATEGORY UCLASS_FPGA +#include <common.h> /* core U-Boot definitions */ #include <console.h> +#include <log.h> #include <ACEX1K.h> /* ACEX device family */ #include <linux/delay.h> -/* Define FPGA_DEBUG to get debug printf's */ -#ifdef FPGA_DEBUG -#define PRINTF(fmt,args...) printf (fmt ,##args) -#else -#define PRINTF(fmt,args...) -#endif -/* Note: The assumption is that we cannot possibly run fast enough to * overrun the device (the Slave Parallel mode can free run at 50MHz). * If there is a need to operate slower, define CONFIG_FPGA_DELAY in
@@ -44,7 +40,7 @@ int ACEX1K_load(Altera_desc *desc, const void *buf, size_t bsize)switch (desc->iface) {case passive_serial: -PRINTF ("%s: Launching Passive Serial Loader\n" , __FUNCTION__); +log_debug ("Launching Passive Serial Loader\n" ); ret_val =ACEX1K_ps_load (desc, buf, bsize);break ;
@@ -64,7 +60,7 @@ int ACEX1K_dump(Altera_desc *desc, const void *buf, size_t bsize)switch (desc->iface) {case passive_serial: -PRINTF ("%s: Launching Passive Serial Dump\n" , __FUNCTION__); +log_debug ("Launching Passive Serial Dump\n" ); ret_val =ACEX1K_ps_dump (desc, buf, bsize);break ;
@@ -93,8 +89,7 @@ static int ACEX1K_ps_load(Altera_desc *desc, const void *buf, size_t bsize)Altera_ACEX1K_Passive_Serial_fns *fn = desc->iface_fns;int i; -PRINTF ("%s: start with interface functions @ 0x%p\n" , - __FUNCTION__, fn); +log_debug ("start with interface functions @ 0x%p\n" , fn);if (fn) {size_t bytecount =0 ;
@@ -102,16 +97,16 @@ static int ACEX1K_ps_load(Altera_desc *desc, const void *buf, size_t bsize)int cookie = desc->cookie;/* make a local copy */ unsigned long ts;/* timestamp */ -PRINTF ("%s: Function Table:\n" -"ptr:\t0x%p\n" -"struct: 0x%p\n" -"config:\t0x%p\n" -"status:\t0x%p\n" -"clk:\t0x%p\n" -"data:\t0x%p\n" -"done:\t0x%p\n\n" , - __FUNCTION__, &fn, fn, fn->config, fn->status, - fn->clk, fn->data, fn->done); +log_debug ("Function Table:\n" +"ptr:\t0x%p\n" +"struct: 0x%p\n" +"config:\t0x%p\n" +"status:\t0x%p\n" +"clk:\t0x%p\n" +"data:\t0x%p\n" +"done:\t0x%p\n\n" , + &fn, fn, fn->config, fn->status, + fn->clk, fn->data, fn->done);#ifdef CONFIG_SYS_FPGA_PROG_FEEDBACK printf ("Loading FPGA Device %d..." , cookie);#endif
Signature · 1 line
2.30.2
PATCH v4 8/10 fpga: spartan2: Use logging feature instead of FPGA_DEBUG
In reply to Alexander DahlInstead of using DEBUG or LOG_DEBUG the driver still had its own definition for debug output.
--- drivers/fpga/spartan2.c | 80 +++++++++++++++++++---------------------- 1 file changed, 37 insertions(+), 43 deletions(-)
drivers/fpga/spartan2.c +37 -43
@@ -4,16 +4,12 @@* Rich Ireland, Enterasys Networks, [email protected]. */ +#define LOG_CATEGORY UCLASS_FPGA +#include <common.h> /* core U-Boot definitions */ +#include <log.h> #include <spartan2.h> /* Spartan-II device family */ -/* Define FPGA_DEBUG to get debug printf's */ -#ifdef FPGA_DEBUG -#define PRINTF(fmt,args...) printf (fmt ,##args) -#else -#define PRINTF(fmt,args...) -#endif -#undef CONFIG_SYS_FPGA_CHECK_BUSY/* Note: The assumption is that we cannot possibly run fast enough to
@@ -46,12 +42,12 @@ static int spartan2_load(xilinx_desc *desc, const void *buf, size_t bsize,switch (desc->iface) {case slave_serial: -PRINTF ("%s: Launching Slave Serial Load\n" , __FUNCTION__); +log_debug ("Launching Slave Serial Load\n" ); ret_val =spartan2_ss_load (desc, buf, bsize);break ;case slave_parallel: -PRINTF ("%s: Launching Slave Parallel Load\n" , __FUNCTION__); +log_debug ("Launching Slave Parallel Load\n" ); ret_val =spartan2_sp_load (desc, buf, bsize);break ;
@@ -69,12 +65,12 @@ static int spartan2_dump(xilinx_desc *desc, const void *buf, size_t bsize)switch (desc->iface) {case slave_serial: -PRINTF ("%s: Launching Slave Serial Dump\n" , __FUNCTION__); +log_debug ("Launching Slave Serial Dump\n" ); ret_val =spartan2_ss_dump (desc, buf, bsize);break ;case slave_parallel: -PRINTF ("%s: Launching Slave Parallel Dump\n" , __FUNCTION__); +log_debug ("Launching Slave Parallel Dump\n" ); ret_val =spartan2_sp_dump (desc, buf, bsize);break ;
@@ -100,8 +96,7 @@ static int spartan2_sp_load(xilinx_desc *desc, const void *buf, size_t bsize)int ret_val =FPGA_FAIL ;/* assume the worst */ xilinx_spartan2_slave_parallel_fns *fn = desc->iface_fns; -PRINTF ("%s: start with interface functions @ 0x%p\n" , - __FUNCTION__, fn); +log_debug ("start with interface functions @ 0x%p\n" , fn);if (fn) {size_t bytecount =0 ;
@@ -109,24 +104,24 @@ static int spartan2_sp_load(xilinx_desc *desc, const void *buf, size_t bsize)int cookie = desc->cookie;/* make a local copy */ unsigned long ts;/* timestamp */ -PRINTF ("%s: Function Table:\n" -"ptr:\t0x%p\n" -"struct: 0x%p\n" -"pre: 0x%p\n" -"pgm:\t0x%p\n" -"init:\t0x%p\n" -"err:\t0x%p\n" -"clk:\t0x%p\n" -"cs:\t0x%p\n" -"wr:\t0x%p\n" -"read data:\t0x%p\n" -"write data:\t0x%p\n" -"busy:\t0x%p\n" -"abort:\t0x%p\n" -"post:\t0x%p\n\n" , - __FUNCTION__, &fn, fn, fn->pre, fn->pgm, fn->init, fn->err, - fn->clk, fn->cs, fn->wr, fn->rdata, fn->wdata, fn->busy, - fn->abort, fn->post); +log_debug ("Function Table:\n" +"ptr:\t0x%p\n" +"struct: 0x%p\n" +"pre: 0x%p\n" +"pgm:\t0x%p\n" +"init:\t0x%p\n" +"err:\t0x%p\n" +"clk:\t0x%p\n" +"cs:\t0x%p\n" +"wr:\t0x%p\n" +"read data:\t0x%p\n" +"write data:\t0x%p\n" +"busy:\t0x%p\n" +"abort:\t0x%p\n" +"post:\t0x%p\n\n" , + &fn, fn, fn->pre, fn->pgm, fn->init, fn->err, + fn->clk, fn->cs, fn->wr, fn->rdata, fn->wdata, fn->busy, + fn->abort, fn->post);/* * This code is designed to emulate the "Express Style"
@@ -302,8 +297,7 @@ static int spartan2_ss_load(xilinx_desc *desc, const void *buf, size_t bsize)int i;unsigned char val; -PRINTF ("%s: start with interface functions @ 0x%p\n" , - __FUNCTION__, fn); +log_debug ("start with interface functions @ 0x%p\n" , fn);if (fn) {size_t bytecount =0 ;
@@ -311,16 +305,16 @@ static int spartan2_ss_load(xilinx_desc *desc, const void *buf, size_t bsize)int cookie = desc->cookie;/* make a local copy */ unsigned long ts;/* timestamp */ -PRINTF ("%s: Function Table:\n" -"ptr:\t0x%p\n" -"struct: 0x%p\n" -"pgm:\t0x%p\n" -"init:\t0x%p\n" -"clk:\t0x%p\n" -"wr:\t0x%p\n" -"done:\t0x%p\n\n" , - __FUNCTION__, &fn, fn, fn->pgm, fn->init, - fn->clk, fn->wr, fn->done); +log_debug ("Function Table:\n" +"ptr:\t0x%p\n" +"struct: 0x%p\n" +"pgm:\t0x%p\n" +"init:\t0x%p\n" +"clk:\t0x%p\n" +"wr:\t0x%p\n" +"done:\t0x%p\n\n" , + &fn, fn, fn->pgm, fn->init, + fn->clk, fn->wr, fn->done);#ifdef CONFIG_SYS_FPGA_PROG_FEEDBACK printf ("Loading FPGA Device %d...\n" , cookie);#endif
Signature · 1 line
2.30.2
PATCH v4 9/10 fpga: spartan3: Use logging feature instead of FPGA_DEBUG
In reply to Alexander DahlInstead of using DEBUG or LOG_DEBUG the driver still had its own definition for debug output.
--- drivers/fpga/spartan3.c | 80 +++++++++++++++++++---------------------- 1 file changed, 37 insertions(+), 43 deletions(-)
drivers/fpga/spartan3.c +37 -43
@@ -9,16 +9,12 @@* on spartan2.c (Rich Ireland, [email protected]). */ +#define LOG_CATEGORY UCLASS_FPGA +#include <common.h> /* core U-Boot definitions */ +#include <log.h> #include <spartan3.h> /* Spartan-II device family */ -/* Define FPGA_DEBUG to get debug printf's */ -#ifdef FPGA_DEBUG -#define PRINTF(fmt,args...) printf (fmt ,##args) -#else -#define PRINTF(fmt,args...) -#endif -#undef CONFIG_SYS_FPGA_CHECK_BUSY/* Note: The assumption is that we cannot possibly run fast enough to
@@ -51,12 +47,12 @@ static int spartan3_load(xilinx_desc *desc, const void *buf, size_t bsize,switch (desc->iface) {case slave_serial: -PRINTF ("%s: Launching Slave Serial Load\n" , __FUNCTION__); +log_debug ("Launching Slave Serial Load\n" ); ret_val =spartan3_ss_load (desc, buf, bsize);break ;case slave_parallel: -PRINTF ("%s: Launching Slave Parallel Load\n" , __FUNCTION__); +log_debug ("Launching Slave Parallel Load\n" ); ret_val =spartan3_sp_load (desc, buf, bsize);break ;
@@ -74,12 +70,12 @@ static int spartan3_dump(xilinx_desc *desc, const void *buf, size_t bsize)switch (desc->iface) {case slave_serial: -PRINTF ("%s: Launching Slave Serial Dump\n" , __FUNCTION__); +log_debug ("Launching Slave Serial Dump\n" ); ret_val =spartan3_ss_dump (desc, buf, bsize);break ;case slave_parallel: -PRINTF ("%s: Launching Slave Parallel Dump\n" , __FUNCTION__); +log_debug ("Launching Slave Parallel Dump\n" ); ret_val =spartan3_sp_dump (desc, buf, bsize);break ;
@@ -105,8 +101,7 @@ static int spartan3_sp_load(xilinx_desc *desc, const void *buf, size_t bsize)int ret_val =FPGA_FAIL ;/* assume the worst */ xilinx_spartan3_slave_parallel_fns *fn = desc->iface_fns; -PRINTF ("%s: start with interface functions @ 0x%p\n" , - __FUNCTION__, fn); +log_debug ("start with interface functions @ 0x%p\n" , fn);if (fn) {size_t bytecount =0 ;
@@ -114,24 +109,24 @@ static int spartan3_sp_load(xilinx_desc *desc, const void *buf, size_t bsize)int cookie = desc->cookie;/* make a local copy */ unsigned long ts;/* timestamp */ -PRINTF ("%s: Function Table:\n" -"ptr:\t0x%p\n" -"struct: 0x%p\n" -"pre: 0x%p\n" -"pgm:\t0x%p\n" -"init:\t0x%p\n" -"err:\t0x%p\n" -"clk:\t0x%p\n" -"cs:\t0x%p\n" -"wr:\t0x%p\n" -"read data:\t0x%p\n" -"write data:\t0x%p\n" -"busy:\t0x%p\n" -"abort:\t0x%p\n" -"post:\t0x%p\n\n" , - __FUNCTION__, &fn, fn, fn->pre, fn->pgm, fn->init, fn->err, - fn->clk, fn->cs, fn->wr, fn->rdata, fn->wdata, fn->busy, - fn->abort, fn->post); +log_debug ("Function Table:\n" +"ptr:\t0x%p\n" +"struct: 0x%p\n" +"pre: 0x%p\n" +"pgm:\t0x%p\n" +"init:\t0x%p\n" +"err:\t0x%p\n" +"clk:\t0x%p\n" +"cs:\t0x%p\n" +"wr:\t0x%p\n" +"read data:\t0x%p\n" +"write data:\t0x%p\n" +"busy:\t0x%p\n" +"abort:\t0x%p\n" +"post:\t0x%p\n\n" , + &fn, fn, fn->pre, fn->pgm, fn->init, fn->err, + fn->clk, fn->cs, fn->wr, fn->rdata, fn->wdata, fn->busy, + fn->abort, fn->post);/* * This code is designed to emulate the "Express Style"
@@ -309,8 +304,7 @@ static int spartan3_ss_load(xilinx_desc *desc, const void *buf, size_t bsize)int i;unsigned char val; -PRINTF ("%s: start with interface functions @ 0x%p\n" , - __FUNCTION__, fn); +log_debug ("start with interface functions @ 0x%p\n" , fn);if (fn) {size_t bytecount =0 ;
@@ -318,16 +312,16 @@ static int spartan3_ss_load(xilinx_desc *desc, const void *buf, size_t bsize)int cookie = desc->cookie;/* make a local copy */ unsigned long ts;/* timestamp */ -PRINTF ("%s: Function Table:\n" -"ptr:\t0x%p\n" -"struct: 0x%p\n" -"pgm:\t0x%p\n" -"init:\t0x%p\n" -"clk:\t0x%p\n" -"wr:\t0x%p\n" -"done:\t0x%p\n\n" , - __FUNCTION__, &fn, fn, fn->pgm, fn->init, - fn->clk, fn->wr, fn->done); +log_debug ("Function Table:\n" +"ptr:\t0x%p\n" +"struct: 0x%p\n" +"pgm:\t0x%p\n" +"init:\t0x%p\n" +"clk:\t0x%p\n" +"wr:\t0x%p\n" +"done:\t0x%p\n\n" , + &fn, fn, fn->pgm, fn->init, + fn->clk, fn->wr, fn->done);#ifdef CONFIG_SYS_FPGA_PROG_FEEDBACK printf ("Loading FPGA Device %d...\n" , cookie);#endif
Signature · 1 line
2.30.2
PATCH v4 10/10 fpga: virtex2: Use logging feature instead of FPGA_DEBUG
In reply to Alexander DahlInstead of using DEBUG or LOG_DEBUG the driver still had its own definition for debug output.
--- drivers/fpga/virtex2.c | 69 ++++++++++++++++++------------------------ 1 file changed, 30 insertions(+), 39 deletions(-)
drivers/fpga/virtex2.c +30 -39
@@ -12,21 +12,14 @@* on spartan2.c (Rich Ireland, [email protected]). */ +#define LOG_CATEGORY UCLASS_FPGA +#include <common.h> #include <console.h> +#include <log.h> #include <virtex2.h> #include <linux/delay.h> -#if 0 -#define FPGA_DEBUG -#endif - -#ifdef FPGA_DEBUG -#define PRINTF(fmt, args...) printf(fmt, ##args) -#else -#define PRINTF(fmt, args...) -#endif -/* * If the SelectMap interface can be overrun by the processor, define * CONFIG_SYS_FPGA_CHECK_BUSY and/or CONFIG_FPGA_DELAY in the board
@@ -89,12 +82,12 @@ static int virtex2_load(xilinx_desc *desc, const void *buf, size_t bsize,switch (desc->iface) {case slave_serial: -PRINTF ("%s: Launching Slave Serial Load\n" , __func__); +log_debug ("Launching Slave Serial Load\n" ); ret_val =virtex2_ss_load (desc, buf, bsize);break ;case slave_selectmap: -PRINTF ("%s: Launching Slave Parallel Load\n" , __func__); +log_debug ("Launching Slave Parallel Load\n" ); ret_val =virtex2_ssm_load (desc, buf, bsize);break ;
@@ -111,12 +104,12 @@ static int virtex2_dump(xilinx_desc *desc, const void *buf, size_t bsize)switch (desc->iface) {case slave_serial: -PRINTF ("%s: Launching Slave Serial Dump\n" , __func__); +log_debug ("Launching Slave Serial Dump\n" ); ret_val =virtex2_ss_dump (desc, buf, bsize);break ;case slave_parallel: -PRINTF ("%s: Launching Slave Parallel Dump\n" , __func__); +log_debug ("Launching Slave Parallel Dump\n" ); ret_val =virtex2_ssm_dump (desc, buf, bsize);break ;
@@ -150,8 +143,7 @@ static int virtex2_slave_pre(xilinx_virtex2_slave_fns *fn, int cookie) {unsigned long ts; -PRINTF ("%s:%d: Start with interface functions @ 0x%p\n" , - __func__, __LINE__, fn); +log_debug ("Start with interface functions @ 0x%p\n" , fn);if (!fn) { printf("%s:%d: NULL Interface function table!\n",
@@ -160,25 +152,24 @@ static int virtex2_slave_pre(xilinx_virtex2_slave_fns *fn, int cookie) }/* Gotta split this one up (so the stack won't blow??) */ -PRINTF ("%s:%d: Function Table:\n" -" base 0x%p\n" -" struct 0x%p\n" -" pre 0x%p\n" -" prog 0x%p\n" -" init 0x%p\n" -" error 0x%p\n" , - __func__, __LINE__, - &fn, fn, fn->pre, fn->pgm, fn->init, fn->err); -PRINTF (" clock 0x%p\n" -" cs 0x%p\n" -" write 0x%p\n" -" rdata 0x%p\n" -" wdata 0x%p\n" -" busy 0x%p\n" -" abort 0x%p\n" -" post 0x%p\n\n" , - fn->clk, fn->cs, fn->wr, fn->rdata, fn->wdata, - fn->busy, fn->abort, fn->post); +log_debug ("Function Table:\n" +" base 0x%p\n" +" struct 0x%p\n" +" pre 0x%p\n" +" prog 0x%p\n" +" init 0x%p\n" +" error 0x%p\n" , + &fn, fn, fn->pre, fn->pgm, fn->init, fn->err); +log_debug (" clock 0x%p\n" +" cs 0x%p\n" +" write 0x%p\n" +" rdata 0x%p\n" +" wdata 0x%p\n" +" busy 0x%p\n" +" abort 0x%p\n" +" post 0x%p\n\n" , + fn->clk, fn->cs, fn->wr, fn->rdata, fn->wdata, + fn->busy, fn->abort, fn->post);#ifdef CONFIG_SYS_FPGA_PROG_FEEDBACK printf ("Initializing FPGA Device %d...\n" , cookie);
@@ -330,8 +321,8 @@ static int virtex2_ssm_load(xilinx_desc *desc, const void *buf, size_t bsize)#endif if ((*fn->done)(cookie) ==FPGA_SUCCESS ) { -PRINTF ("%s:%d:done went active early, bytecount = %zu\n" , - __func__, __LINE__, bytecount); +log_debug ("done went active early, bytecount = %zu\n" , + bytecount);break ; }
@@ -465,8 +456,8 @@ static int virtex2_ss_load(xilinx_desc *desc, const void *buf, size_t bsize)#endif if ((*fn->done)(cookie) ==FPGA_SUCCESS ) { -PRINTF ("%s:%d:done went active early, bytecount = %zu\n" , - __func__, __LINE__, bytecount); +log_debug ("done went active early, bytecount = %zu\n" , + bytecount);break ; }
Signature · 1 line
2.30.2
Michal Simek
AppliedQuoted 68 lines · Alexander Dahl
Hei hei,
while working on FPGA support for a new device I discovered debug logging in some FPGA drivers is still done as in the old days. Bring that to what I thougt would be the currently preferred approach.
Notes: Adding those
Kconfigsymbols in patch 1 is just to be able to build two of the old drivers.All drivers touched were build tested with
sandbox_defconfigand GCC 8 on Debian GNU/Linux 10 (buster).Lines with other possibly questionable output were not touched, only what seemed to be designated debug output, and only for FPGA drivers having that ancient
FPGA_DEBUG/ PRINTF macros, so there's room for future improvements.Changelog:
v3 -> v4:
- Reordered patches,Kconfigpatch comes first now (made it easier to build and test the series step by step)
- Added three patches fixing printf compiler warnings first, before changing to the new logging framework (so CI should not fail anymore with -Werror)v2 -> v3:
- Patch introducing FPGA uclass was completely reworked, sent independently from this series, and applied already, thus removed
- Because requiring that new FPGA uclass changes, rebased on Michal's microblaze branch '20221005'
- Removed '"%s …",__func__' and '"%d …",__line__' from log messages, because log framework can add those (enabled byCONFIG_LOGF_FUNCandCONFIG_LOGF_LINE)v1 -> v2:
- Rebased on master
- Added patch to introduce new FPGA uclass in front of the other patches
- Use that new uclass as log category
- Slightly reworded cover letterGreets
AlexCc Michal Simek <[email protected]>Alexander Dahl (10):
fpga: Add missingKconfigsymbols for old FPGA drivers
fpga: spartan2: Fix printf arguments warning
fpga: spartan3: Fix printf arguments warning
fpga: virtex2: Fix printf format string warnings
fpga: altera: Use logging feature instead ofFPGA_DEBUG
fpga: cyclon2: Use logging feature instead ofFPGA_DEBUG
fpga: ACEX1K: Use logging feature instead ofFPGA_DEBUG
fpga: spartan2: Use logging feature instead ofFPGA_DEBUG
fpga: spartan3: Use logging feature instead ofFPGA_DEBUG
fpga: virtex2: Use logging feature instead ofFPGA_DEBUGdrivers/fpga/ACEX1K.c | 37 +++++++++---------- drivers/fpga/Kconfig | 12 +++++++ drivers/fpga/altera.c | 11 +++--- drivers/fpga/cyclon2.c | 38 +++++++++----------- drivers/fpga/spartan2.c | 80 +++++++++++++++++++---------------------- drivers/fpga/spartan3.c | 80 +++++++++++++++++++---------------------- drivers/fpga/virtex2.c | 69 ++++++++++++++++------------------- 7 files changed, 152 insertions(+), 175 deletions(-)
Quoted 1 line
base-commit:
2d8cf392a77815f062446ef441f1078958dc1b2a
Applied.
M