[U-Boot] [PATCH] 0/12 Multiadapter/multibus I2C
Here is the second attempt for initial portion of multibus/multiadapter I2C support.
This includes a set of common files, all drivers in drivers/i2c and all boards affected by these changes (config files, board files, and lib_xx files.)
There is an illustrative example of multiadapter multibus I2C config in MPC8548CDS.h config file (#if 0'd.) Definitions in that example are bogus so please don't expect it to work. It will compile though...
This set also includes big rework for soft_i2c.c that makes it template version that allows up to 4 bitbanged adapters. This number can be reduced/increased; 4 is arbitrary chosen value. I'm not a CPP guru so I did not find a way to make a version without a limit. Sure I could add additional configuration variable for this limit but it is one more variable and resulting soft_i2c.c would be much uglier.
The general rule for multi-adapter controllers (fsl_i2c, mxc_i2c, etc.) is that all defines for the first controller do not have any number in their names, the second one has "2" and so on. That allows for compatibility with existing code. All existing defaults are kept.
I did test-compile all ARM and PPC boards. Unfortunately I have never worked with m68k, mips, blackfin so I don't have toolchains for those to do test builds.
The only board that failed to compile was SIMPC8313.h that fails to link NAND bootstrap with "NAND bootstrap too big" but it has nothing to do with new I2C code; it also fails without my patches.
All patches are against today's u-boot-i2c.git.
--- ****************************************************************** * KSI@home KOI8 Net < > The impossible we do immediately. * * Las Vegas NV, USA < > Miracles require 24-hour notice. * ******************************************************************
Hello ksi,
ksi@koi8.net wrote:
Here is the second attempt for initial portion of multibus/multiadapter I2C support.
Can you please send your patches with some better commit messages. You only send your Signed-off-by, without any explanation. Please change this.
This includes a set of common files, all drivers in drivers/i2c and all boards affected by these changes (config files, board files, and lib_xx files.)
There is an illustrative example of multiadapter multibus I2C config in MPC8548CDS.h config file (#if 0'd.) Definitions in that example are bogus so please don't expect it to work. It will compile though...
This set also includes big rework for soft_i2c.c that makes it template version that allows up to 4 bitbanged adapters. This number can be
Didn;t you try my suggestion? This is a really big define monster now, which I think, we can avoid, and without to change nearly all lines of the existing driver.
reduced/increased; 4 is arbitrary chosen value. I'm not a CPP guru so I did not find a way to make a version without a limit. Sure I could add additional configuration variable for this limit but it is one more variable and resulting soft_i2c.c would be much uglier.
The general rule for multi-adapter controllers (fsl_i2c, mxc_i2c, etc.) is that all defines for the first controller do not have any number in their names, the second one has "2" and so on. That allows for compatibility with existing code. All existing defaults are kept.
I did test-compile all ARM and PPC boards. Unfortunately I have never worked with m68k, mips, blackfin so I don't have toolchains for those to do test builds.
The only board that failed to compile was SIMPC8313.h that fails to link NAND bootstrap with "NAND bootstrap too big" but it has nothing to do with new I2C code; it also fails without my patches.
All patches are against today's u-boot-i2c.git.
bye Heiko
On Fri, 13 Feb 2009, Heiko Schocher wrote:
Hello ksi,
ksi@koi8.net wrote:
Here is the second attempt for initial portion of multibus/multiadapter I2C support.
Can you please send your patches with some better commit messages. You only send your Signed-off-by, without any explanation. Please change this.
There is not much sense in extensive commit messages in this case, IMHO. It is not a bug fix or added feature at one particular place; it is a major rework. The only message I can give is something like "Changes for multiadapter/multibus I2C support..."
I'll add it to the second attempt that I will make later today.
This includes a set of common files, all drivers in drivers/i2c and all boards affected by these changes (config files, board files, and lib_xx files.)
There is an illustrative example of multiadapter multibus I2C config in MPC8548CDS.h config file (#if 0'd.) Definitions in that example are bogus so please don't expect it to work. It will compile though...
This set also includes big rework for soft_i2c.c that makes it template version that allows up to 4 bitbanged adapters. This number can be
Didn;t you try my suggestion? This is a really big define monster now, which I think, we can avoid, and without to change nearly all lines of the existing driver.
We can not avoid it. At least I can not see an easy way to do this. SOFT_I2C is special. Those multiple e.g. MXC or OMAP3 adapters can be parameterized because different channels do only differ in their base address that can be made into a parameter. Software I2C is totally different because it has totally different functions for different channels, there is nothing we can make into a parameter. All those I2C_SDA etc. are NOT DEFINES, they are MACROS. Every function for every channel is built of those macros that can be absolutely different for each channel. They define NOT some PARAMETERS but function TEXT that will be compiled into executable code.
This is how it is done in Linux kernel (see e.g. drivers/hwmon/lm93.c.)
--- ****************************************************************** * KSI@home KOI8 Net < > The impossible we do immediately. * * Las Vegas NV, USA < > Miracles require 24-hour notice. * ******************************************************************
Hello ksi,
ksi@koi8.net wrote:
On Fri, 13 Feb 2009, Heiko Schocher wrote:
ksi@koi8.net wrote:
Here is the second attempt for initial portion of multibus/multiadapter I2C support.
Can you please send your patches with some better commit messages. You only send your Signed-off-by, without any explanation. Please change this.
There is not much sense in extensive commit messages in this case, IMHO. It is not a bug fix or added feature at one particular place; it is a major rework. The only message I can give is something like "Changes for multiadapter/multibus I2C support..."
I'll add it to the second attempt that I will make later today.
This includes a set of common files, all drivers in drivers/i2c and all boards affected by these changes (config files, board files, and lib_xx files.)
There is an illustrative example of multiadapter multibus I2C config in MPC8548CDS.h config file (#if 0'd.) Definitions in that example are bogus so please don't expect it to work. It will compile though...
This set also includes big rework for soft_i2c.c that makes it template version that allows up to 4 bitbanged adapters. This number can be
Didn;t you try my suggestion? This is a really big define monster now, which I think, we can avoid, and without to change nearly all lines of the existing driver.
We can not avoid it. At least I can not see an easy way to do this. SOFT_I2C
Yes, we can. Look again deeper in my approach, this _is_ an easy way!
I also looked again in your changes in the fsl_i2c.c driver, and we can make this also simplier, by using cur_adap_nr->hwadapnr!
We have not to define for both hardware adapter each function in i2c_adap_t! For example:
You did: static void fsl_i2c1_init(int speed, int slaveadd) { __i2c_init(0, speed, slaveadd); }
instead we only need
i2c_init(cur_adap_nr->hwadapnr, speed, slaveadd);
with
i2c_adap_t fsl_i2c_adap[] = { { .init = i2c_init, [...] .hwadapnr = 0, .name = FSL_NAME(CONFIG_SYS_FSL_I2C_OFFSET) }, #ifdef CONFIG_SYS_FSL_I2C2_OFFSET { .init = i2c_init, [...] .hwadapnr = 1, .name = FSL_NAME(CONFIG_SYS_FSL_I2C2_OFFSET) }, #endif
Please change this driver also!
If I think more, we never even need to change the function parameters like you did for example for i2c_int ()! We can use at the beginning of every function who go in i2c_adap_t, the "cur_adap_nr->hwadapnr" and make the settings we need for this function... and wow we saved one function parameter.
is special. Those multiple e.g. MXC or OMAP3 adapters can be parameterized because different channels do only differ in their base address that can be made into a parameter. Software I2C is totally different because it has
Why is this different? If you change a base or the way to the pins?
totally different functions for different channels, there is nothing we can
Think about my explanation to the soft_i2c.c driver in previous EMail and above function.
It also works!
make into a parameter. All those I2C_SDA etc. are NOT DEFINES, they are MACROS. Every function for every channel is built of those macros that can
I know this in your approach, but we _don;t_ need this. We simply can make one "common" board function and switch in this function dependent on the "cur_adap_nr->hwadapnr" to the particular GPIO pin functions, wherever the are!
be absolutely different for each channel. They define NOT some PARAMETERS but function TEXT that will be compiled into executable code.
And this additional TEXT I save too! I think, you think too much in C++?
This is how it is done in Linux kernel (see e.g. drivers/hwmon/lm93.c.)
Thats no argument.
bye Heiko
On Sat, 14 Feb 2009, Heiko Schocher wrote:
Hello ksi,
ksi@koi8.net wrote:
On Fri, 13 Feb 2009, Heiko Schocher wrote:
ksi@koi8.net wrote:
Here is the second attempt for initial portion of multibus/multiadapter I2C support.
Can you please send your patches with some better commit messages. You only send your Signed-off-by, without any explanation. Please change this.
There is not much sense in extensive commit messages in this case, IMHO. It is not a bug fix or added feature at one particular place; it is a major rework. The only message I can give is something like "Changes for multiadapter/multibus I2C support..."
I'll add it to the second attempt that I will make later today.
This includes a set of common files, all drivers in drivers/i2c and all boards affected by these changes (config files, board files, and lib_xx files.)
There is an illustrative example of multiadapter multibus I2C config in MPC8548CDS.h config file (#if 0'd.) Definitions in that example are bogus so please don't expect it to work. It will compile though...
This set also includes big rework for soft_i2c.c that makes it template version that allows up to 4 bitbanged adapters. This number can be
Didn;t you try my suggestion? This is a really big define monster now, which I think, we can avoid, and without to change nearly all lines of the existing driver.
We can not avoid it. At least I can not see an easy way to do this. SOFT_I2C
Yes, we can. Look again deeper in my approach, this _is_ an easy way!
I also looked again in your changes in the fsl_i2c.c driver, and we can make this also simplier, by using cur_adap_nr->hwadapnr!
OK, please explain how that cur_adap_nr->hwadapnr gets assigned. Please also explain how can one invoke a function on other adapter than "current". Remember, i2c_init is quite often called BEFORE the code is relocated to RAM so you can NOT change "current" adapter.
Please also note that you will loose a capability of working with more than one adapter before the code is relocated to RAM.
We have not to define for both hardware adapter each function in i2c_adap_t! For example:
You did: static void fsl_i2c1_init(int speed, int slaveadd) { __i2c_init(0, speed, slaveadd); }
instead we only need
i2c_init(cur_adap_nr->hwadapnr, speed, slaveadd);
with
i2c_adap_t fsl_i2c_adap[] = { { .init = i2c_init, [...] .hwadapnr = 0, .name = FSL_NAME(CONFIG_SYS_FSL_I2C_OFFSET) }, #ifdef CONFIG_SYS_FSL_I2C2_OFFSET { .init = i2c_init, [...] .hwadapnr = 1, .name = FSL_NAME(CONFIG_SYS_FSL_I2C2_OFFSET) }, #endif
It would've been easy if we had had "this" pointer. That would allow us to find out what adapter we are running on by using something like "this->hwadapnr." Unfortunately we do NOT have such a pointer, we're plain C. Function in a structure does not have a way to find out how to access a member of that structure. The only way to somehow find which "hwadapnr" we are running at is using a global variable, cur_i2c_bus as a starting point. But that is meaningless until the code is relocated to RAM and that variable became writable. And that robs us of added possibility of using any adapter other than a single one preset in config file before relocating to RAM.
That is if we want to keep the original I2C API. The other, simpler way is to add an argument to each and every function, a pointer to i2c_adap_t structure or its index or something similar. But that defeats the entire purpose of this code by requiring to find and change each and every call to any I2C function in the entire U-Boot source thus totally breaking ALL existing code 99.99% of which only use single I2C adapter/bus...
Please change this driver also!
I can't. Please read above.
If I think more, we never even need to change the function parameters like you did for example for i2c_int ()! We can use at the beginning of every function who go in i2c_adap_t, the "cur_adap_nr->hwadapnr" and make the settings we need for this function... and wow we saved one function parameter.
Devil is in the details... Please read above.
is special. Those multiple e.g. MXC or OMAP3 adapters can be parameterized because different channels do only differ in their base address that can be made into a parameter. Software I2C is totally different because it has
Why is this different? If you change a base or the way to the pins?
Because the pins on different channels can be accessesed in absolutely different way.
totally different functions for different channels, there is nothing we can
Think about my explanation to the soft_i2c.c driver in previous EMail and above function.
It also works!
Partially and with handicaps. Please read my reply to that message.
make into a parameter. All those I2C_SDA etc. are NOT DEFINES, they are MACROS. Every function for every channel is built of those macros that can
I know this in your approach, but we _don;t_ need this. We simply can make one "common" board function and switch in this function dependent on the "cur_adap_nr->hwadapnr" to the particular GPIO pin functions, wherever the are!
Please read above.
be absolutely different for each channel. They define NOT some PARAMETERS but function TEXT that will be compiled into executable code.
And this additional TEXT I save too!
You don't save anything. And you add complexity and break uniformity. BTW, what is a reason to save on text?
I think, you think too much in C++?
No, I'm not a C++ guy. I prefer assembly and only use C where it is absolutely necessary :) I do not program in C++/Java/etc. at all and only know C++ basics and a bit of other stuff required to fix occasional bugs in other people's code.
It is not C++, it is common sence and good engineering practices. And I'm an engineer first and only then a programmer, a machinist, an electronic engineer, a gunsmith and so on.
This is how it is done in Linux kernel (see e.g. drivers/hwmon/lm93.c.)
Thats no argument.
Sure it is not. This is a suggestion.
--- ****************************************************************** * KSI@home KOI8 Net < > The impossible we do immediately. * * Las Vegas NV, USA < > Miracles require 24-hour notice. * ******************************************************************
Hello ksi,
ksi@koi8.net wrote:
On Sat, 14 Feb 2009, Heiko Schocher wrote:
Hello ksi,
ksi@koi8.net wrote:
On Fri, 13 Feb 2009, Heiko Schocher wrote:
ksi@koi8.net wrote:
Here is the second attempt for initial portion of multibus/multiadapter I2C support.
Can you please send your patches with some better commit messages. You only send your Signed-off-by, without any explanation. Please change this.
There is not much sense in extensive commit messages in this case, IMHO. It is not a bug fix or added feature at one particular place; it is a major rework. The only message I can give is something like "Changes for multiadapter/multibus I2C support..."
I'll add it to the second attempt that I will make later today.
This includes a set of common files, all drivers in drivers/i2c and all boards affected by these changes (config files, board files, and lib_xx files.)
There is an illustrative example of multiadapter multibus I2C config in MPC8548CDS.h config file (#if 0'd.) Definitions in that example are bogus so please don't expect it to work. It will compile though...
This set also includes big rework for soft_i2c.c that makes it template version that allows up to 4 bitbanged adapters. This number can be
Didn;t you try my suggestion? This is a really big define monster now, which I think, we can avoid, and without to change nearly all lines of the existing driver.
We can not avoid it. At least I can not see an easy way to do this. SOFT_I2C
Yes, we can. Look again deeper in my approach, this _is_ an easy way!
I also looked again in your changes in the fsl_i2c.c driver, and we can make this also simplier, by using cur_adap_nr->hwadapnr!
OK, please explain how that cur_adap_nr->hwadapnr gets assigned. Please also
When running from ram, this is no problem. It should be set in i2c_set_bus_num().
explain how can one invoke a function on other adapter than "current".
Is this needed? If so, you must before call a i2c_set_bus_num(), and after you finished call it again with the old busnumber. So it is done for example in do_date () common/cmd_date.c
Remember, i2c_init is quite often called BEFORE the code is relocated to RAM so you can NOT change "current" adapter.
Yes, thats a point. But do we need this before running from ram (except one hardwareadapter)?
Please also note that you will loose a capability of working with more than one adapter before the code is relocated to RAM.
Do we really need this?
We have not to define for both hardware adapter each function in i2c_adap_t! For example:
You did: static void fsl_i2c1_init(int speed, int slaveadd) { __i2c_init(0, speed, slaveadd); }
instead we only need
i2c_init(cur_adap_nr->hwadapnr, speed, slaveadd);
with
i2c_adap_t fsl_i2c_adap[] = { { .init = i2c_init, [...] .hwadapnr = 0, .name = FSL_NAME(CONFIG_SYS_FSL_I2C_OFFSET) }, #ifdef CONFIG_SYS_FSL_I2C2_OFFSET { .init = i2c_init, [...] .hwadapnr = 1, .name = FSL_NAME(CONFIG_SYS_FSL_I2C2_OFFSET) }, #endif
It would've been easy if we had had "this" pointer. That would allow us to find out what adapter we are running on by using something like "this->hwadapnr." Unfortunately we do NOT have such a pointer, we're plain C. Function in a structure does not have a way to find out how to access a member of that structure. The only way to somehow find which "hwadapnr" we are running at is using a global variable, cur_i2c_bus as a starting point. But that is meaningless until the code is relocated to RAM and that variable became writable. And that robs us of added possibility of using any adapter other than a single one preset in config file before relocating to RAM.
Yes, I know. But again, do we need this?
That is if we want to keep the original I2C API. The other, simpler way is to add an argument to each and every function, a pointer to i2c_adap_t structure or its index or something similar. But that defeats the entire purpose of this code by requiring to find and change each and every call to any I2C function in the entire U-Boot source thus totally breaking ALL existing code 99.99% of which only use single I2C adapter/bus...
That would be a hard way.
Please change this driver also!
I can't. Please read above.
If I think more, we never even need to change the function parameters like you did for example for i2c_int ()! We can use at the beginning of every function who go in i2c_adap_t, the "cur_adap_nr->hwadapnr" and make the settings we need for this function... and wow we saved one function parameter.
Devil is in the details... Please read above.
Thats why we discuss it ;-)
is special. Those multiple e.g. MXC or OMAP3 adapters can be parameterized because different channels do only differ in their base address that can be made into a parameter. Software I2C is totally different because it has
Why is this different? If you change a base or the way to the pins?
Because the pins on different channels can be accessesed in absolutely different way.
totally different functions for different channels, there is nothing we can
Think about my explanation to the soft_i2c.c driver in previous EMail and above function.
It also works!
Partially and with handicaps. Please read my reply to that message.
If we really need more then one bus when running from flash, this is a problem.
make into a parameter. All those I2C_SDA etc. are NOT DEFINES, they are MACROS. Every function for every channel is built of those macros that can
I know this in your approach, but we _don;t_ need this. We simply can make one "common" board function and switch in this function dependent on the "cur_adap_nr->hwadapnr" to the particular GPIO pin functions, wherever the are!
Please read above.
be absolutely different for each channel. They define NOT some PARAMETERS but function TEXT that will be compiled into executable code.
And this additional TEXT I save too!
You don't save anything. And you add complexity and break uniformity. BTW,
I save text when having 4 bitbang drivers running. And I don;t see where it is complexer nor where it breaks uniformity.
what is a reason to save on text?
We are only a bootloader and have often to fit in a maybe small flash.
bye Heiko
On Sun, 15 Feb 2009, Heiko Schocher wrote:
Hello ksi,
ksi@koi8.net wrote:
On Sat, 14 Feb 2009, Heiko Schocher wrote:
Hello ksi,
ksi@koi8.net wrote:
On Fri, 13 Feb 2009, Heiko Schocher wrote:
ksi@koi8.net wrote:
Here is the second attempt for initial portion of multibus/multiadapter I2C support.
Can you please send your patches with some better commit messages. You only send your Signed-off-by, without any explanation. Please change this.
There is not much sense in extensive commit messages in this case, IMHO. It is not a bug fix or added feature at one particular place; it is a major rework. The only message I can give is something like "Changes for multiadapter/multibus I2C support..."
I'll add it to the second attempt that I will make later today.
This includes a set of common files, all drivers in drivers/i2c and all boards affected by these changes (config files, board files, and lib_xx files.)
There is an illustrative example of multiadapter multibus I2C config in MPC8548CDS.h config file (#if 0'd.) Definitions in that example are bogus so please don't expect it to work. It will compile though...
This set also includes big rework for soft_i2c.c that makes it template version that allows up to 4 bitbanged adapters. This number can be
Didn;t you try my suggestion? This is a really big define monster now, which I think, we can avoid, and without to change nearly all lines of the existing driver.
We can not avoid it. At least I can not see an easy way to do this. SOFT_I2C
Yes, we can. Look again deeper in my approach, this _is_ an easy way!
I also looked again in your changes in the fsl_i2c.c driver, and we can make this also simplier, by using cur_adap_nr->hwadapnr!
OK, please explain how that cur_adap_nr->hwadapnr gets assigned. Please also
When running from ram, this is no problem. It should be set in i2c_set_bus_num().
Yep. But nobody's perfect and you can have a situation when you need to access several busses before relocation. It is not hardware for U-Boot, it is U-Boot for hardware. When hardware designers design their hardware they don't make their decisions based on U-Boot limitation. That is us who should accomodate what they designed.
There is also another consideration -- when having several adapters which one should be initialized at boot time, before relocation? Another problem is init() function that can be unique for each adapter. To make the lower layer transparent I'm reprogramming muxes if any when switching busses. It is necessary to make I2C API simple and uniform between muxed and non-muxed busses. That essentially means that we can NOT do i2c_set_bus_num() to execute init() for a particular adapter -- adapter MUST be initialized for i2c_set_bus_num() to succeed.
Your suggestion requires total LOGIC change.
explain how can one invoke a function on other adapter than "current".
Is this needed? If so, you must before call a i2c_set_bus_num(), and after you finished call it again with the old busnumber. So it is done for example in do_date () common/cmd_date.c
You can not do it before all adapters are initialized. And you WON'T be able to initialize adapters because you will not be able to switch busses.
Remember, i2c_init is quite often called BEFORE the code is relocated to RAM so you can NOT change "current" adapter.
Yes, thats a point. But do we need this before running from ram (except one hardwareadapter)?
Yes, see above.
Please also note that you will loose a capability of working with more than one adapter before the code is relocated to RAM.
Do we really need this?
Eh, "640K ought to be enough to anybody..."
We have not to define for both hardware adapter each function in i2c_adap_t! For example:
You did: static void fsl_i2c1_init(int speed, int slaveadd) { __i2c_init(0, speed, slaveadd); }
instead we only need
i2c_init(cur_adap_nr->hwadapnr, speed, slaveadd);
with
i2c_adap_t fsl_i2c_adap[] = { { .init = i2c_init, [...] .hwadapnr = 0, .name = FSL_NAME(CONFIG_SYS_FSL_I2C_OFFSET) }, #ifdef CONFIG_SYS_FSL_I2C2_OFFSET { .init = i2c_init, [...] .hwadapnr = 1, .name = FSL_NAME(CONFIG_SYS_FSL_I2C2_OFFSET) }, #endif
It would've been easy if we had had "this" pointer. That would allow us to find out what adapter we are running on by using something like "this->hwadapnr." Unfortunately we do NOT have such a pointer, we're plain C. Function in a structure does not have a way to find out how to access a member of that structure. The only way to somehow find which "hwadapnr" we are running at is using a global variable, cur_i2c_bus as a starting point. But that is meaningless until the code is relocated to RAM and that variable became writable. And that robs us of added possibility of using any adapter other than a single one preset in config file before relocating to RAM.
Yes, I know. But again, do we need this?
We do. Otherwise we can essentially throw everything to trash and start over. This requires changing the logical design, architecture. And this is that logic that is most difficult and takes most thinking. Coding is easy.
That is if we want to keep the original I2C API. The other, simpler way is to add an argument to each and every function, a pointer to i2c_adap_t structure or its index or something similar. But that defeats the entire purpose of this code by requiring to find and change each and every call to any I2C function in the entire U-Boot source thus totally breaking ALL existing code 99.99% of which only use single I2C adapter/bus...
That would be a hard way.
That is why I spent a week thinking about the design that would allow to keep most of existing code.
Please change this driver also!
I can't. Please read above.
If I think more, we never even need to change the function parameters like you did for example for i2c_int ()! We can use at the beginning of every function who go in i2c_adap_t, the "cur_adap_nr->hwadapnr" and make the settings we need for this function... and wow we saved one function parameter.
Devil is in the details... Please read above.
Thats why we discuss it ;-)
is special. Those multiple e.g. MXC or OMAP3 adapters can be parameterized because different channels do only differ in their base address that can be made into a parameter. Software I2C is totally different because it has
Why is this different? If you change a base or the way to the pins?
Because the pins on different channels can be accessesed in absolutely different way.
totally different functions for different channels, there is nothing we can
Think about my explanation to the soft_i2c.c driver in previous EMail and above function.
It also works!
Partially and with handicaps. Please read my reply to that message.
If we really need more then one bus when running from flash, this is a problem.
No, that's not just that. There are multiple reasons why the original driver had been made with macros.
First, it is _SMALLER_ when done this way. Most of those macros (I2C_SCL etc.) translate into 1 to 3 assembly instructions depending on particular processor code set. Except some special cases the most complex operation they do is changing a bit at some address that takes 3 instructions if particular CPU can not change set/reset bits directly - read->modify->write. Many CPUs can make it in 1 to 2 instructions.
There is no way how you can avoid those instructions -- the work must be done. You insist on making them into functions (there is no other way if they reside in another object file.) That means that you do NOT eliminate those instructions, you replace them with function calls. Even if we assume that those function calls will not have any prolog/epilog many processors would not be allow you to perform a function call with a single instruction because they do not have e.g. automatic stack so you should save your link register before the call ad restore it after. That already makes _MORE_ code per operation than those macros had. And that is just function calls overhead, the actual function body is extra...
Another reason why macros are used is speed. Not everyone is running U-Boot on 10 GHz Pentium-9 with gigabyte of cache. In bitbanged I2C every instruction counts if you want to run a bus at a decent speed (I won't even start with regular 100kHz less for 400kHz; 50kHz would be very good.) Your approach adds function call overhead to every memory/io bit change because you must perform a function call every time you want to read SDA line or change SCL state etc. It is ON TOP of what those macros were doing because no matter what the real work must be done so that code in those macros still must be executed. And remember, there is probably no instruction cache and you're running off of flash so every instruction fetch is something like 100ns for a single word. That is for a single flash access but you would probably need at least 2 such accesses for a single-word instruction fetch because most of the CPUs are 32-bit and I really doubt anybody uses 32-bit flash for U-Boot... That makes it 400 ns just to fetch _ONE_ that "branch" instruction on 32-bit CPU running off of 8-bit flash. And another 400 ns to fetch a "return" instruction after the real work is done. That is if your CPU has automatic stack. Multiply it by two if it doesn't. That makes it 1.6 us just for a single function call overhead.
And you will NOT save a single byte of program code. It will probably get even bigger because you will have to add that conditional logic for checking what the current bus is and perform appropriate actions based on that.
make into a parameter. All those I2C_SDA etc. are NOT DEFINES, they are MACROS. Every function for every channel is built of those macros that can
I know this in your approach, but we _don;t_ need this. We simply can make one "common" board function and switch in this function dependent on the "cur_adap_nr->hwadapnr" to the particular GPIO pin functions, wherever the are!
Please read above.
be absolutely different for each channel. They define NOT some PARAMETERS but function TEXT that will be compiled into executable code.
And this additional TEXT I save too!
You don't save anything. And you add complexity and break uniformity. BTW,
I save text when having 4 bitbang drivers running. And I don;t see where it is complexer nor where it breaks uniformity.
what is a reason to save on text?
We are only a bootloader and have often to fit in a maybe small flash.
You ain't gonna save anything. Quite in contrary that approach will almost certainly make the code bigger.
--- ****************************************************************** * KSI@home KOI8 Net < > The impossible we do immediately. * * Las Vegas NV, USA < > Miracles require 24-hour notice. * ******************************************************************
Hello ksi,
ksi@koi8.net wrote:
On Sun, 15 Feb 2009, Heiko Schocher wrote:
ksi@koi8.net wrote:
On Sat, 14 Feb 2009, Heiko Schocher wrote:
ksi@koi8.net wrote:
ksi@koi8.net wrote:
> Here is the second attempt for initial portion of multibus/multiadapter > I2C support.
[...]
When running from ram, this is no problem. It should be set in i2c_set_bus_num().
Yep. But nobody's perfect and you can have a situation when you need to access several busses before relocation. It is not hardware for U-Boot, it is U-Boot for hardware. When hardware designers design their hardware they don't make their decisions based on U-Boot limitation. That is us who should accomodate what they designed.
Din;t know, there is such a design.
There is also another consideration -- when having several adapters which one should be initialized at boot time, before relocation? Another problem is init() function that can be unique for each adapter. To make the lower layer transparent I'm reprogramming muxes if any when switching busses. It is necessary to make I2C API simple and uniform between muxed and non-muxed busses. That essentially means that we can NOT do i2c_set_bus_num() to execute init() for a particular adapter -- adapter MUST be initialized for i2c_set_bus_num() to succeed.
Hmm.. okay, but we can also call from i2c_set_bus_num() when running in flash first the init() function ... but that wouldn;t be nice, you are right.
Remember, i2c_init is quite often called BEFORE the code is relocated to RAM so you can NOT change "current" adapter.
Yes, thats a point. But do we need this before running from ram (except one hardwareadapter)?
Yes, see above.
Yes, thats is a problem in my approach, and if we need more then one i2c bus when running from flash, maybe a no go.
[...]
That is if we want to keep the original I2C API. The other, simpler way is to add an argument to each and every function, a pointer to i2c_adap_t structure or its index or something similar. But that defeats the entire purpose of this code by requiring to find and change each and every call to any I2C function in the entire U-Boot source thus totally breaking ALL existing code 99.99% of which only use single I2C adapter/bus...
That would be a hard way.
That is why I spent a week thinking about the design that would allow to keep most of existing code.
Thats why I discuss with you, to get this infos ;-)
[...]
If we really need more then one bus when running from flash, this is a problem.
No, that's not just that. There are multiple reasons why the original driver had been made with macros.
First, it is _SMALLER_ when done this way. Most of those macros (I2C_SCL etc.) translate into 1 to 3 assembly instructions depending on particular processor code set. Except some special cases the most complex operation they do is changing a bit at some address that takes 3 instructions if particular CPU can not change set/reset bits directly - read->modify->write. Many CPUs can make it in 1 to 2 instructions.
Ok.
There is no way how you can avoid those instructions -- the work must be done. You insist on making them into functions (there is no other way if they reside in another object file.) That means that you do NOT eliminate
As I said, it should be possible to do this also in macros. But you are right, there is always a +switch, which will cost more code ...
bye Heiko
Dear Heiko,
In message 49992BEB.30501@denx.de you wrote:
Yes, thats a point. But do we need this before running from ram (except one hardwareadapter)?
Yes, see above.
Yes, thats is a problem in my approach, and if we need more then one i2c bus when running from flash, maybe a no go.
It is not an unsolvable problem. We do have the global data structure, and it has been (mis-) used for worse purposes before.
Best regards,
Wolfgang Denk
On Mon, 16 Feb 2009, Wolfgang Denk wrote:
Dear Heiko,
In message 49992BEB.30501@denx.de you wrote:
Yes, thats a point. But do we need this before running from ram (except one hardwareadapter)?
Yes, see above.
Yes, thats is a problem in my approach, and if we need more then one i2c bus when running from flash, maybe a no go.
It is not an unsolvable problem. We do have the global data structure, and it has been (mis-) used for worse purposes before.
Not so fast... And remember, there are other similar interfaces that are waiting in the line (e.g. SPI...)
I personally do not think we should use that global data structure for such obscure purposes. And it must be writable for this to work.
--- ****************************************************************** * KSI@home KOI8 Net < > The impossible we do immediately. * * Las Vegas NV, USA < > Miracles require 24-hour notice. * ******************************************************************
Dear ksi@koi8.net,
In message Pine.LNX.4.64ksi.0902162152530.27482@home-gw.koi8.net you wrote:
It is not an unsolvable problem. We do have the global data structure, and it has been (mis-) used for worse purposes before.
Not so fast... And remember, there are other similar interfaces that are waiting in the line (e.g. SPI...)
So what?
SPI is an even more exotic interface, with even less need to use it ever before relocation to RAM. Let's not artificially construct new problems when there are none.
I personally do not think we should use that global data structure for such obscure purposes. And it must be writable for this to work.
"obscure purposes"? Hm... The global data structure was created for this very purpose: to provide writable global storage before relocation.
Best regards,
Wolfgang Denk
Dear ksi@koi8.net,
In message Pine.LNX.4.64ksi.0902152236430.17769@home-gw.koi8.net you wrote:
Yep. But nobody's perfect and you can have a situation when you need to access several busses before relocation. It is not hardware for U-Boot, it is U-Boot for hardware. When hardware designers design their hardware they don't make their decisions based on U-Boot limitation. That is us who should accomodate what they designed.
We don't have to make the mainline U-Boot implementation unnecessary complex just for the small chance that there is sombeody stupid enough to design a broken system.
It is our task as software engineers to tell the hardware designers which designs are easy to support, nd which will cause problems in the software.
There is also another consideration -- when having several adapters which one should be initialized at boot time, before relocation? Another problem
The one that is needed there, if any at all.
is init() function that can be unique for each adapter. To make the lower layer transparent I'm reprogramming muxes if any when switching busses. It is necessary to make I2C API simple and uniform between muxed and non-muxed busses. That essentially means that we can NOT do i2c_set_bus_num() to execute init() for a particular adapter -- adapter MUST be initialized for i2c_set_bus_num() to succeed.
Your suggestion requires total LOGIC change.
Maybe. But that's why we're discussing this here.
Is this needed? If so, you must before call a i2c_set_bus_num(), and after you finished call it again with the old busnumber. So it is done for example in do_date () common/cmd_date.c
You can not do it before all adapters are initialized. And you WON'T be able to initialize adapters because you will not be able to switch busses.
This sounds like a design problem to me, then.
Please keep in mind that according to U-Boot philosophy it is forbidden to always initialize all adapters. Only those actually used by U-Boot may be initialized, and shall be de-initialized after use.
Yes, thats a point. But do we need this before running from ram (except one hardwareadapter)?
Yes, see above.
Um... Maybe I missed something - did you give an example (except for broken designs) where this really might be needed?
Yes, I know. But again, do we need this?
We do. Otherwise we can essentially throw everything to trash and start over. This requires changing the logical design, architecture. And this is that logic that is most difficult and takes most thinking. Coding is easy.
You say: "we do [need that]". I ask: why? what for?
Another reason why macros are used is speed. Not everyone is running U-Boot on 10 GHz Pentium-9 with gigabyte of cache. In bitbanged I2C every instruction counts if you want to run a bus at a decent speed (I won't even start with regular 100kHz less for 400kHz; 50kHz would be very good.) Your
Are you sure? If I remember correctly soft-I2C can even run 400 KHz on a slow 50 MHz MPC8xx system.
Do you have other numbers?
must be executed. And remember, there is probably no instruction cache and you're running off of flash so every instruction fetch is something like
I tend to say that an U-Boot port where instruction cache is disabled is misconfigured and should be fixed :-)
Best regards,
Wolfgang Denk
On Mon, 16 Feb 2009, Wolfgang Denk wrote:
Dear ksi@koi8.net,
In message Pine.LNX.4.64ksi.0902152236430.17769@home-gw.koi8.net you wrote:
Yep. But nobody's perfect and you can have a situation when you need to access several busses before relocation. It is not hardware for U-Boot, it is U-Boot for hardware. When hardware designers design their hardware they don't make their decisions based on U-Boot limitation. That is us who should accomodate what they designed.
We don't have to make the mainline U-Boot implementation unnecessary complex just for the small chance that there is sombeody stupid enough to design a broken system.
It is our task as software engineers to tell the hardware designers which designs are easy to support, nd which will cause problems in the software.
It is not complex. Quite in the contrary, it is much simpler and more straightforward to have 2 sets of functions for 2 adapters in one place than to have one that includes both of them in one place switchable using some global variable at another place used by wrappers in a third place.
There is also another consideration -- when having several adapters which one should be initialized at boot time, before relocation? Another problem
The one that is needed there, if any at all.
You will have to pick that one for each and every board somehow...
is init() function that can be unique for each adapter. To make the lower layer transparent I'm reprogramming muxes if any when switching busses. It is necessary to make I2C API simple and uniform between muxed and non-muxed busses. That essentially means that we can NOT do i2c_set_bus_num() to execute init() for a particular adapter -- adapter MUST be initialized for i2c_set_bus_num() to succeed.
Your suggestion requires total LOGIC change.
Maybe. But that's why we're discussing this here.
Is this needed? If so, you must before call a i2c_set_bus_num(), and after you finished call it again with the old busnumber. So it is done for example in do_date () common/cmd_date.c
You can not do it before all adapters are initialized. And you WON'T be able to initialize adapters because you will not be able to switch busses.
This sounds like a design problem to me, then.
Please keep in mind that according to U-Boot philosophy it is forbidden to always initialize all adapters. Only those actually used by U-Boot may be initialized, and shall be de-initialized after use.
Eh, I don't see any problem initializing all I2C controllers at the same time. What is the problem with that?
Then, we _DO_ already initialize _ALL_ controllers in U-Boot. Most of the time the total number of controllers is 1 and we do initialize all of them. In those rare cases when that number is not 1, we _DO_ initialize all of them (look at fsl_i2c.c as it is in the main tree right now.) Existing i2c_init() function does not have any provision for initializing any particular controller; it initializes them all.
Yes, thats a point. But do we need this before running from ram (except one hardwareadapter)?
Yes, see above.
Um... Maybe I missed something - did you give an example (except for broken designs) where this really might be needed?
We did not have such boards as of now but that does not mean they can't exist. I would've agreed for making a handicapped version if it had been difficult to make a full-featured one but it is not. Furthermore, additional efforts required to put such a handicap on existing code.
Yes, I know. But again, do we need this?
We do. Otherwise we can essentially throw everything to trash and start over. This requires changing the logical design, architecture. And this is that logic that is most difficult and takes most thinking. Coding is easy.
You say: "we do [need that]". I ask: why? what for?
Another reason why macros are used is speed. Not everyone is running U-Boot on 10 GHz Pentium-9 with gigabyte of cache. In bitbanged I2C every instruction counts if you want to run a bus at a decent speed (I won't even start with regular 100kHz less for 400kHz; 50kHz would be very good.) Your
Are you sure? If I remember correctly soft-I2C can even run 400 KHz on a slow 50 MHz MPC8xx system.
Do you have other numbers?
Is something wrong with those numbers?
must be executed. And remember, there is probably no instruction cache and you're running off of flash so every instruction fetch is something like
I tend to say that an U-Boot port where instruction cache is disabled is misconfigured and should be fixed :-)
Eh, everything is easy when you have your RAM running and all handicaps dropped off... Just try to run 100 meters with heavy weights on your legs and you will see you can not compete with sprinters :)
--- ****************************************************************** * KSI@home KOI8 Net < > The impossible we do immediately. * * Las Vegas NV, USA < > Miracles require 24-hour notice. * ******************************************************************
Dear ksi@koi8.net,
In message Pine.LNX.4.64ksi.0902162132540.27482@home-gw.koi8.net you wrote:
It is our task as software engineers to tell the hardware designers which designs are easy to support, nd which will cause problems in the software.
It is not complex. Quite in the contrary, it is much simpler and more straightforward to have 2 sets of functions for 2 adapters in one place than to have one that includes both of them in one place switchable using some global variable at another place used by wrappers in a third place.
To code I've seen so far looked pretty complex, and the fact that we've spend many rounds of discussion here seems to indicate a certain level of cmplexity, too.
There is also another consideration -- when having several adapters which one should be initialized at boot time, before relocation? Another problem
The one that is needed there, if any at all.
You will have to pick that one for each and every board somehow...
Only for those boards that need it, whichis a tiny fraction of all boards which use I2C. And there,it should be trivial to configure in the board config file.
Please keep in mind that according to U-Boot philosophy it is forbidden to always initialize all adapters. Only those actually used by U-Boot may be initialized, and shall be de-initialized after use.
Eh, I don't see any problem initializing all I2C controllers at the same time. What is the problem with that?
Why would you do that? It makes little sense to initialize components that may never be accessed at all.
Then, we _DO_ already initialize _ALL_ controllers in U-Boot. Most of the time the total number of controllers is 1 and we do initialize all of them. In those rare cases when that number is not 1, we _DO_ initialize all of them (look at fsl_i2c.c as it is in the main tree right now.) Existing i2c_init() function does not have any provision for initializing any particular controller; it initializes them all.
When we re-implementing that code, we should fix that error instead of copying it.
Another reason why macros are used is speed. Not everyone is running U-Boot on 10 GHz Pentium-9 with gigabyte of cache. In bitbanged I2C every instruction counts if you want to run a bus at a decent speed (I won't even start with regular 100kHz less for 400kHz; 50kHz would be very good.) Your
Are you sure? If I remember correctly soft-I2C can even run 400 KHz on a slow 50 MHz MPC8xx system.
Do you have other numbers?
Is something wrong with those numbers?
You seem to claim that soft-I2C cannot run at 400 kHz clock. I doubt that claim.
I tend to say that an U-Boot port where instruction cache is disabled is misconfigured and should be fixed :-)
Eh, everything is easy when you have your RAM running and all handicaps dropped off... Just try to run 100 meters with heavy weights on your legs and you will see you can not compete with sprinters :)
We're at the point: who needs I2C before relocation to RAM? Only a small fraction of all boards. How many of these use soft-I2C? An even smaller fraction. For how many of these is it critical to run at 100 kHz I2C clock or faster? Well, I think the answer is none here.
We're wasting time and efforts on a non-issue.
Best regards,
Wolfgang Denk
Dear ksi@koi8.net,
In message Pine.LNX.4.64ksi.0902142104100.6240@home-gw.koi8.net you wrote:
OK, please explain how that cur_adap_nr->hwadapnr gets assigned. Please also explain how can one invoke a function on other adapter than "current". Remember, i2c_init is quite often called BEFORE the code is relocated to RAM so you can NOT change "current" adapter.
We could assign an entry in the global data for it.
But then - how often will it bbe necessary to switch adapters before relocation? What is I2C being used for? To read the SPD data for the RAM init code. Which other adapter would be needed?
Please also note that you will loose a capability of working with more than one adapter before the code is relocated to RAM.
I don't see this.
Best regards,
Wolfgang Denk
On Mon, 16 Feb 2009, Wolfgang Denk wrote:
Dear ksi@koi8.net,
In message Pine.LNX.4.64ksi.0902142104100.6240@home-gw.koi8.net you wrote:
OK, please explain how that cur_adap_nr->hwadapnr gets assigned. Please also explain how can one invoke a function on other adapter than "current". Remember, i2c_init is quite often called BEFORE the code is relocated to RAM so you can NOT change "current" adapter.
We could assign an entry in the global data for it.
But then - how often will it bbe necessary to switch adapters before relocation? What is I2C being used for? To read the SPD data for the RAM init code. Which other adapter would be needed?
I can not tell it right away. There might be several RAM DIMMs with their own SPD EPROMS sitting on different busses or something else. It is not like we absolutely need this feature right now, urgently but it is so easy to implement that it would've been a crime to not to... :)
There is another problem with choosing a proper i2c_init() function with i2c_set_bus_num() -- the latter also reprograms I2C multiplexers/switches if there are any so you essentially need an initialized adapter to switch to it to initialize it. Catch 22...
Please also note that you will loose a capability of working with more than one adapter before the code is relocated to RAM.
I don't see this.
Why? There is no writable global variables before relocation...
--- ****************************************************************** * KSI@home KOI8 Net < > The impossible we do immediately. * * Las Vegas NV, USA < > Miracles require 24-hour notice. * ******************************************************************
Hello ksi,
ksi@koi8.net wrote:
On Mon, 16 Feb 2009, Wolfgang Denk wrote:
Dear ksi@koi8.net,
In message Pine.LNX.4.64ksi.0902142104100.6240@home-gw.koi8.net you wrote:
OK, please explain how that cur_adap_nr->hwadapnr gets assigned. Please also explain how can one invoke a function on other adapter than "current". Remember, i2c_init is quite often called BEFORE the code is relocated to RAM so you can NOT change "current" adapter.
We could assign an entry in the global data for it.
But then - how often will it bbe necessary to switch adapters before relocation? What is I2C being used for? To read the SPD data for the RAM init code. Which other adapter would be needed?
I can not tell it right away. There might be several RAM DIMMs with their own SPD EPROMS sitting on different busses or something else. It is not like we absolutely need this feature right now, urgently but it is so easy to implement that it would've been a crime to not to... :)
There is another problem with choosing a proper i2c_init() function with i2c_set_bus_num() -- the latter also reprograms I2C multiplexers/switches if there are any so you essentially need an initialized adapter to switch to it to initialize it. Catch 22...
That would not be a problem:
if we are running from flash we could always init in i2c_set_bus_num() first the new controllor to which we switch. And if we are running in ram we can always check in i2c_set_bus_num() if adap->init_done in i2c_adap_t is 1. If this is not so, we first init the bus! With this, it should be possible to get rid of all this i2c_init calls all over the code ...
One requirement for this: All "old" code who directly call i2c_read, i2c_write, ... must, if this new feature is activated, call a i2c_set_bus_num() before accessing the bus. But this would be only "work" (grep and fix).
bye
Heiko
Dear ksi@koi8.net,
In message Pine.LNX.4.64ksi.0902162123560.27482@home-gw.koi8.net you wrote:
But then - how often will it bbe necessary to switch adapters before relocation? What is I2C being used for? To read the SPD data for the RAM init code. Which other adapter would be needed?
I can not tell it right away. There might be several RAM DIMMs with their own SPD EPROMS sitting on different busses or something else. It is not like
I have never seen any such design in real life, and I would not hesitate to call it broken. We should not make the software design unnecessarily complet just to support any worst case situations that might eventually happen. It's perfectly sufficient to cover 99.999% of all existing systems :-)
we absolutely need this feature right now, urgently but it is so easy to implement that it would've been a crime to not to... :)
Well, it's not so easy - it adds a lot of complexity, it seems.
There is another problem with choosing a proper i2c_init() function with i2c_set_bus_num() -- the latter also reprograms I2C multiplexers/switches if there are any so you essentially need an initialized adapter to switch to it to initialize it. Catch 22...
One more reason not even to attempt to support such configurations.
Please also note that you will loose a capability of working with more than one adapter before the code is relocated to RAM.
I don't see this.
Why? There is no writable global variables before relocation...
Of course there are. Guess what the "global data" strcuture has been invented for?
Best regards,
Wolfgang Denk
Dear ksi@koi8.net,
In message Pine.LNX.4.64ksi.0902131156250.26866@home-gw.koi8.net you wrote:
Can you please send your patches with some better commit messages. You only send your Signed-off-by, without any explanation. Please change this.
There is not much sense in extensive commit messages in this case, IMHO. It is not a bug fix or added feature at one particular place; it is a major rework. The only message I can give is something like "Changes for multiadapter/multibus I2C support..."
This sounds very much as if there was a danger that the patches are not logically cut - did you make sure that eahc of the commits is still bisectable?
Best regards,
Wolfgang Denk
On Mon, 16 Feb 2009, Wolfgang Denk wrote:
Dear ksi@koi8.net,
In message Pine.LNX.4.64ksi.0902131156250.26866@home-gw.koi8.net you wrote:
Can you please send your patches with some better commit messages. You only send your Signed-off-by, without any explanation. Please change this.
There is not much sense in extensive commit messages in this case, IMHO. It is not a bug fix or added feature at one particular place; it is a major rework. The only message I can give is something like "Changes for multiadapter/multibus I2C support..."
This sounds very much as if there was a danger that the patches are not logically cut - did you make sure that eahc of the commits is still bisectable?
I will rebase them and slice 'em in smaller pieces.
Unfortunately this is a major rework so there is no way one can make small patches really independent. Changes to include/i2c.h, e.g. make the entire old code uncompilable so they only make sence together with some other parts that constitute the core of the new code. Everything else is much easier when the core is in.
--- ****************************************************************** * KSI@home KOI8 Net < > The impossible we do immediately. * * Las Vegas NV, USA < > Miracles require 24-hour notice. * ******************************************************************
participants (3)
-
Heiko Schocher -
ksi＠koi8.net -
Wolfgang Denk