Re: [U-Boot] [PATCH V3 1/5] pxa: move i2c driver to the common place
> -Original Message- > From: Heiko Schocher [mailto:h...@denx.de] > Sent: Wednesday, March 23, 2011 2:37 PM > To: Lei Wen > Cc: Prafulla Wadaskar; Lei Wen; Wolfgang Denk; u-boot@lists.denx.de; > Marek Vasut; Ashish Karkare; Prabhanjan Sarnaik; Yu Tang > Subject: Re: [PATCH V3 1/5] pxa: move i2c driver to the common place > > Hello Lei, > > Lei Wen wrote: > > Hi Heiko, > > > > On Wed, Mar 23, 2011 at 4:53 PM, Heiko Schocher wrote: > >> Hello Lei, > >> > >> Lei Wen wrote: > >>> Hi Heiko, > >>> > >>> On Wed, Mar 23, 2011 at 4:22 PM, Heiko Schocher wrote: > Hello Prafulla, > > Prafulla Wadaskar wrote: > >> -Original Message- > >> From: Lei Wen [mailto:adrian.w...@gmail.com] > >> Sent: Tuesday, March 22, 2011 6:14 PM > >> To: Prafulla Wadaskar > >> Cc: Lei Wen; Heiko Schocher; Wolfgang Denk; u-boot@lists.denx.de; > Marek > >> Vasut; Ashish Karkare; Prabhanjan Sarnaik; Yu Tang > >> Subject: Re: [PATCH V3 1/5] pxa: move i2c driver to the common > place > >> > [...] > >>> Also my incoming mmc/sd enabling patch for pantheon and armada100 > >>> is also based on this i2c enabling patch, as I need the i2c to turn > on the > >>> repsonding pmic power connection. > >>> > >>> While we could get a i2c working pantheon, armada100, and other pxa > >>> series platform now with this patch set. So what about Could we > merge > >>> this first, and > >>> gradually change to the i2c framework, test and make it mature. > >> I am fine with that, but please address the other comments from > >> Prafulla and Wolfgang (rename defines, use standard accessors). > > > > Ok, I would post another round of patch to fix those issue Prafulla > and Wolfgang > > metioned. > > Ok, thanks! > > >> If you plan to investigate time for the multibus/multiadapter > >> i2c branch, let me know, maybe I can rebase this branch before > >> you use it. > > > > Currently, I have no bandwidth to do this... :( > > Maybe I would try to investigate when I have time. > > :-( > > Thats also my state ... please give me a notice, if you find some time, > so we can sync us ... Sorry I posted earlier email before checking this email. Even same the my state too :-) Well Let's lei get his stuff, we will think of migration latter :-( Regards.. Prafulla . . ___ U-Boot mailing list U-Boot@lists.denx.de http://lists.denx.de/mailman/listinfo/u-boot
Re: [U-Boot] [PATCH V3 1/5] pxa: move i2c driver to the common place
> -Original Message- > From: Heiko Schocher [mailto:h...@denx.de] > Sent: Wednesday, March 23, 2011 2:23 PM > To: Lei Wen > Cc: Prafulla Wadaskar; Lei Wen; Wolfgang Denk; u-boot@lists.denx.de; > Marek Vasut; Ashish Karkare; Prabhanjan Sarnaik; Yu Tang > Subject: Re: [PATCH V3 1/5] pxa: move i2c driver to the common place > > Hello Lei, > > Lei Wen wrote: > > Hi Heiko, > > > > On Wed, Mar 23, 2011 at 4:22 PM, Heiko Schocher wrote: > >> Hello Prafulla, > >> > >> Prafulla Wadaskar wrote: > -Original Message- > From: Lei Wen [mailto:adrian.w...@gmail.com] > Sent: Tuesday, March 22, 2011 6:14 PM > To: Prafulla Wadaskar > Cc: Lei Wen; Heiko Schocher; Wolfgang Denk; u-boot@lists.denx.de; > Marek > Vasut; Ashish Karkare; Prabhanjan Sarnaik; Yu Tang > Subject: Re: [PATCH V3 1/5] pxa: move i2c driver to the common > place > > >>> ...snip... > >> drivers/i2c/Makefile |1 + > >> drivers/i2c/mv_i2c.c | 452 > >> +++ > >> include/configs/innokom.h |1 + > >> include/configs/xm250.h |1 + > >> 6 files changed, 455 insertions(+), 470 deletions(-) > >> delete mode 100644 arch/arm/cpu/pxa/i2c.c > >> create mode 100644 drivers/i2c/mv_i2c.c > > ...snip... > > > >> -#endif /* CONFIG_HARD_I2C */ > >> diff --git a/drivers/i2c/Makefile b/drivers/i2c/Makefile > >> index 052fe36..00a12cc 100644 > >> --- a/drivers/i2c/Makefile > >> +++ b/drivers/i2c/Makefile > >> @@ -29,6 +29,7 @@ COBJS-$(CONFIG_BFIN_TWI_I2C) += bfin-twi_i2c.o > >> COBJS-$(CONFIG_DRIVER_DAVINCI_I2C) += davinci_i2c.o > >> COBJS-$(CONFIG_FSL_I2C) += fsl_i2c.o > >> COBJS-$(CONFIG_I2C_MVTWSI) += mvtwsi.o > >> +COBJS-$(CONFIG_I2C_MV) += mv_i2c.o > > Mvtwsi and mv_i2c are two i2c drivers for Marvell. > > Can you merge these two? > As I explain to you before. Although kirkwood and pxa series are > both > the product > of Marvell, but it don't necessary means that they must have the > same > controller > for both product line. For the i2c part, they just use two > different > controller. > So why you keep request merge those two? Do you mean you want to > create a unique I2C > framework for whole i2c drivers in drivers/i2c? > >>> Hi Lei > >>> > >>> 1. Most of i2c drivers supported in u-boot are either in SoC > specific folder or in drivers/i2c folder, there is no as such thumb rule > here. > >> New drivers should go to drivers/i2c ! > >> The existing (old) drivers are just not moved ... > >> patches welcome! > >> > >>> 2. Secondly all these drivers have some common code, mostly > i2c_read, i2c_write, i2c_probe, etc.. > >>> 3. Specific to Marvell, we already have mvtwsi.c that supports > Kirkwood and Orion5X SoCs. Whereas you are adding new mvi2c.c that will > support armada100, pantheon apart from pxa. > >>> 4. What about if we need to support some new Marvell SoC with > different i2C controller? Do we add one more driver? > >>> > >>> I would love if some one creates drivers/i2c/i2c_core.c??? not > necessarily you ;-) > >>> > >>> Here is what I would like to suggest. > >>> 1. cmd_i2c mostly interfaced with i2c_probe, i2c_read, i2c_write, > i2c_get_bun_num, i2c_set_bus_num, those should go in drivers/i2c_core.c > >> Yep, but see below comment. > >> > >>> 2. APIs like i2c_start, i2c_stop, i2c_send, i2c_recv, i2c_reset are > more I2C controller specific and those will be different implementation > on different SoCs, those can go in SoC specific i2c driver file. > >> Yep. > >> > >>> 3. all I2C driver files should be in drivers/i2c/ > >> Yep. > >> > >>> 4. i2c_read/write API need to be redefined since those are not > generic to be used to access any I2C peripheral( most of the device > don't need address to be programmed) > >> With which devices do you have problems? You can set with > >> "i2c mw chip address.0 ..." an addresslen = 0 ... or? > >> > >>> 5. Flags must be provided for i2c_read/write APIs to have precise > control to execute I2C_START/I2C_STOP sequence in the call. > >> If needed, yes. > >> > >>> Since you are the one starting with re-using pxa driver for > armada100 and Pantheon SoC, why don't you split it into i2c_core.c and > i2c_pxa.c? then add i2c_armada100.c and i2c_pantheon.c? > >>> Others can migrate in the similar way. (even mvtwsi,c) > >>> > >>> Hi Heiko > >>> What do you think on this? > >> I made such a i2c_core.c file in the multibus/multiadapter branch > >> for the i2c subsystem, see here: > >> > >> http://git.denx.de/?p=u-boot/u-boot- > i2c.git;a=shortlog;h=refs/heads/multibus_v2 > >> > >> (also you can grep in u-boot ML for discussions about) > >> > >> Actual state: > >> - arm boards: i2c driver tested on suen3 (kirkwood based board) > >> - powerpc boards: i2c driver tested on mpc82xx, 83xx, 8xx boards > >> - all others just coded not tested ... > >> (A result of lacking hw and/or ti
Re: [U-Boot] [PATCH V3 1/5] pxa: move i2c driver to the common place
Hello Lei, Lei Wen wrote: > Hi Heiko, > > On Wed, Mar 23, 2011 at 4:53 PM, Heiko Schocher wrote: >> Hello Lei, >> >> Lei Wen wrote: >>> Hi Heiko, >>> >>> On Wed, Mar 23, 2011 at 4:22 PM, Heiko Schocher wrote: Hello Prafulla, Prafulla Wadaskar wrote: >> -Original Message- >> From: Lei Wen [mailto:adrian.w...@gmail.com] >> Sent: Tuesday, March 22, 2011 6:14 PM >> To: Prafulla Wadaskar >> Cc: Lei Wen; Heiko Schocher; Wolfgang Denk; u-boot@lists.denx.de; Marek >> Vasut; Ashish Karkare; Prabhanjan Sarnaik; Yu Tang >> Subject: Re: [PATCH V3 1/5] pxa: move i2c driver to the common place >> [...] >>> Also my incoming mmc/sd enabling patch for pantheon and armada100 >>> is also based on this i2c enabling patch, as I need the i2c to turn on the >>> repsonding pmic power connection. >>> >>> While we could get a i2c working pantheon, armada100, and other pxa >>> series platform now with this patch set. So what about Could we merge >>> this first, and >>> gradually change to the i2c framework, test and make it mature. >> I am fine with that, but please address the other comments from >> Prafulla and Wolfgang (rename defines, use standard accessors). > > Ok, I would post another round of patch to fix those issue Prafulla and > Wolfgang > metioned. Ok, thanks! >> If you plan to investigate time for the multibus/multiadapter >> i2c branch, let me know, maybe I can rebase this branch before >> you use it. > > Currently, I have no bandwidth to do this... :( > Maybe I would try to investigate when I have time. :-( Thats also my state ... please give me a notice, if you find some time, so we can sync us ... bye, Heiko -- DENX Software Engineering GmbH, MD: Wolfgang Denk & Detlev Zundel HRB 165235 Munich, Office: Kirchenstr.5, D-82194 Groebenzell, Germany ___ U-Boot mailing list U-Boot@lists.denx.de http://lists.denx.de/mailman/listinfo/u-boot
Re: [U-Boot] [PATCH V3 1/5] pxa: move i2c driver to the common place
Hi Heiko, On Wed, Mar 23, 2011 at 4:53 PM, Heiko Schocher wrote: > Hello Lei, > > Lei Wen wrote: >> Hi Heiko, >> >> On Wed, Mar 23, 2011 at 4:22 PM, Heiko Schocher wrote: >>> Hello Prafulla, >>> >>> Prafulla Wadaskar wrote: > -Original Message- > From: Lei Wen [mailto:adrian.w...@gmail.com] > Sent: Tuesday, March 22, 2011 6:14 PM > To: Prafulla Wadaskar > Cc: Lei Wen; Heiko Schocher; Wolfgang Denk; u-boot@lists.denx.de; Marek > Vasut; Ashish Karkare; Prabhanjan Sarnaik; Yu Tang > Subject: Re: [PATCH V3 1/5] pxa: move i2c driver to the common place > ...snip... >>> drivers/i2c/Makefile | 1 + >>> drivers/i2c/mv_i2c.c | 452 >>> +++ >>> include/configs/innokom.h | 1 + >>> include/configs/xm250.h | 1 + >>> 6 files changed, 455 insertions(+), 470 deletions(-) >>> delete mode 100644 arch/arm/cpu/pxa/i2c.c >>> create mode 100644 drivers/i2c/mv_i2c.c >> ...snip... >> >>> -#endif /* CONFIG_HARD_I2C */ >>> diff --git a/drivers/i2c/Makefile b/drivers/i2c/Makefile >>> index 052fe36..00a12cc 100644 >>> --- a/drivers/i2c/Makefile >>> +++ b/drivers/i2c/Makefile >>> @@ -29,6 +29,7 @@ COBJS-$(CONFIG_BFIN_TWI_I2C) += bfin-twi_i2c.o >>> COBJS-$(CONFIG_DRIVER_DAVINCI_I2C) += davinci_i2c.o >>> COBJS-$(CONFIG_FSL_I2C) += fsl_i2c.o >>> COBJS-$(CONFIG_I2C_MVTWSI) += mvtwsi.o >>> +COBJS-$(CONFIG_I2C_MV) += mv_i2c.o >> Mvtwsi and mv_i2c are two i2c drivers for Marvell. >> Can you merge these two? > As I explain to you before. Although kirkwood and pxa series are both > the product > of Marvell, but it don't necessary means that they must have the same > controller > for both product line. For the i2c part, they just use two different > controller. > So why you keep request merge those two? Do you mean you want to > create a unique I2C > framework for whole i2c drivers in drivers/i2c? Hi Lei 1. Most of i2c drivers supported in u-boot are either in SoC specific folder or in drivers/i2c folder, there is no as such thumb rule here. >>> New drivers should go to drivers/i2c ! >>> The existing (old) drivers are just not moved ... >>> patches welcome! >>> 2. Secondly all these drivers have some common code, mostly i2c_read, i2c_write, i2c_probe, etc.. 3. Specific to Marvell, we already have mvtwsi.c that supports Kirkwood and Orion5X SoCs. Whereas you are adding new mvi2c.c that will support armada100, pantheon apart from pxa. 4. What about if we need to support some new Marvell SoC with different i2C controller? Do we add one more driver? I would love if some one creates drivers/i2c/i2c_core.c??? not necessarily you ;-) Here is what I would like to suggest. 1. cmd_i2c mostly interfaced with i2c_probe, i2c_read, i2c_write, i2c_get_bun_num, i2c_set_bus_num, those should go in drivers/i2c_core.c >>> Yep, but see below comment. >>> 2. APIs like i2c_start, i2c_stop, i2c_send, i2c_recv, i2c_reset are more I2C controller specific and those will be different implementation on different SoCs, those can go in SoC specific i2c driver file. >>> Yep. >>> 3. all I2C driver files should be in drivers/i2c/ >>> Yep. >>> 4. i2c_read/write API need to be redefined since those are not generic to be used to access any I2C peripheral( most of the device don't need address to be programmed) >>> With which devices do you have problems? You can set with >>> "i2c mw chip address.0 ..." an addresslen = 0 ... or? >>> 5. Flags must be provided for i2c_read/write APIs to have precise control to execute I2C_START/I2C_STOP sequence in the call. >>> If needed, yes. >>> Since you are the one starting with re-using pxa driver for armada100 and Pantheon SoC, why don't you split it into i2c_core.c and i2c_pxa.c? then add i2c_armada100.c and i2c_pantheon.c? Others can migrate in the similar way. (even mvtwsi,c) Hi Heiko What do you think on this? >>> I made such a i2c_core.c file in the multibus/multiadapter branch >>> for the i2c subsystem, see here: >>> >>> http://git.denx.de/?p=u-boot/u-boot-i2c.git;a=shortlog;h=refs/heads/multibus_v2 >>> >>> (also you can grep in u-boot ML for discussions about) >>> >>> Actual state: >>> - arm boards: i2c driver tested on suen3 (kirkwood based board) >>> - powerpc boards: i2c driver tested on mpc82xx, 83xx, 8xx boards >>> - all others just coded not tested ... >>> (A result of lacking hw and/or time) >>> >>> ToDo: >>> - rebase against current head >>> (Sorry, didn;t found time to rebase it since Oktober 2010) >>> - Update README >>> - porting arrived new i2c drivers, boards since Oktober 2010 >>> to this new i2c approach >>> - testing, testing, testing ... Testers welcome! >>> >>> I prefer to integrate this to
Re: [U-Boot] [PATCH V3 1/5] pxa: move i2c driver to the common place
Hello Lei, Lei Wen wrote: > Hi Heiko, > > On Wed, Mar 23, 2011 at 4:22 PM, Heiko Schocher wrote: >> Hello Prafulla, >> >> Prafulla Wadaskar wrote: -Original Message- From: Lei Wen [mailto:adrian.w...@gmail.com] Sent: Tuesday, March 22, 2011 6:14 PM To: Prafulla Wadaskar Cc: Lei Wen; Heiko Schocher; Wolfgang Denk; u-boot@lists.denx.de; Marek Vasut; Ashish Karkare; Prabhanjan Sarnaik; Yu Tang Subject: Re: [PATCH V3 1/5] pxa: move i2c driver to the common place >>> ...snip... >> drivers/i2c/Makefile |1 + >> drivers/i2c/mv_i2c.c | 452 >> +++ >> include/configs/innokom.h |1 + >> include/configs/xm250.h |1 + >> 6 files changed, 455 insertions(+), 470 deletions(-) >> delete mode 100644 arch/arm/cpu/pxa/i2c.c >> create mode 100644 drivers/i2c/mv_i2c.c > ...snip... > >> -#endif /* CONFIG_HARD_I2C */ >> diff --git a/drivers/i2c/Makefile b/drivers/i2c/Makefile >> index 052fe36..00a12cc 100644 >> --- a/drivers/i2c/Makefile >> +++ b/drivers/i2c/Makefile >> @@ -29,6 +29,7 @@ COBJS-$(CONFIG_BFIN_TWI_I2C) += bfin-twi_i2c.o >> COBJS-$(CONFIG_DRIVER_DAVINCI_I2C) += davinci_i2c.o >> COBJS-$(CONFIG_FSL_I2C) += fsl_i2c.o >> COBJS-$(CONFIG_I2C_MVTWSI) += mvtwsi.o >> +COBJS-$(CONFIG_I2C_MV) += mv_i2c.o > Mvtwsi and mv_i2c are two i2c drivers for Marvell. > Can you merge these two? As I explain to you before. Although kirkwood and pxa series are both the product of Marvell, but it don't necessary means that they must have the same controller for both product line. For the i2c part, they just use two different controller. So why you keep request merge those two? Do you mean you want to create a unique I2C framework for whole i2c drivers in drivers/i2c? >>> Hi Lei >>> >>> 1. Most of i2c drivers supported in u-boot are either in SoC specific >>> folder or in drivers/i2c folder, there is no as such thumb rule here. >> New drivers should go to drivers/i2c ! >> The existing (old) drivers are just not moved ... >> patches welcome! >> >>> 2. Secondly all these drivers have some common code, mostly i2c_read, >>> i2c_write, i2c_probe, etc.. >>> 3. Specific to Marvell, we already have mvtwsi.c that supports Kirkwood and >>> Orion5X SoCs. Whereas you are adding new mvi2c.c that will support >>> armada100, pantheon apart from pxa. >>> 4. What about if we need to support some new Marvell SoC with different i2C >>> controller? Do we add one more driver? >>> >>> I would love if some one creates drivers/i2c/i2c_core.c??? not necessarily >>> you ;-) >>> >>> Here is what I would like to suggest. >>> 1. cmd_i2c mostly interfaced with i2c_probe, i2c_read, i2c_write, >>> i2c_get_bun_num, i2c_set_bus_num, those should go in drivers/i2c_core.c >> Yep, but see below comment. >> >>> 2. APIs like i2c_start, i2c_stop, i2c_send, i2c_recv, i2c_reset are more >>> I2C controller specific and those will be different implementation on >>> different SoCs, those can go in SoC specific i2c driver file. >> Yep. >> >>> 3. all I2C driver files should be in drivers/i2c/ >> Yep. >> >>> 4. i2c_read/write API need to be redefined since those are not generic to >>> be used to access any I2C peripheral( most of the device don't need address >>> to be programmed) >> With which devices do you have problems? You can set with >> "i2c mw chip address.0 ..." an addresslen = 0 ... or? >> >>> 5. Flags must be provided for i2c_read/write APIs to have precise control >>> to execute I2C_START/I2C_STOP sequence in the call. >> If needed, yes. >> >>> Since you are the one starting with re-using pxa driver for armada100 and >>> Pantheon SoC, why don't you split it into i2c_core.c and i2c_pxa.c? then >>> add i2c_armada100.c and i2c_pantheon.c? >>> Others can migrate in the similar way. (even mvtwsi,c) >>> >>> Hi Heiko >>> What do you think on this? >> I made such a i2c_core.c file in the multibus/multiadapter branch >> for the i2c subsystem, see here: >> >> http://git.denx.de/?p=u-boot/u-boot-i2c.git;a=shortlog;h=refs/heads/multibus_v2 >> >> (also you can grep in u-boot ML for discussions about) >> >> Actual state: >> - arm boards: i2c driver tested on suen3 (kirkwood based board) >> - powerpc boards: i2c driver tested on mpc82xx, 83xx, 8xx boards >> - all others just coded not tested ... >> (A result of lacking hw and/or time) >> >> ToDo: >> - rebase against current head >> (Sorry, didn;t found time to rebase it since Oktober 2010) >> - Update README >> - porting arrived new i2c drivers, boards since Oktober 2010 >> to this new i2c approach >> - testing, testing, testing ... Testers welcome! >> >> I prefer to integrate this to mainline, before we do above steps >> (4?) and 5. As Lei mentioned, if a soc/board has different i2c >> controllers and more than one bus we *need* this approach, >> so it is not worthwhil
Re: [U-Boot] [PATCH V3 1/5] pxa: move i2c driver to the common place
Hi Heiko, On Wed, Mar 23, 2011 at 4:22 PM, Heiko Schocher wrote: > Hello Prafulla, > > Prafulla Wadaskar wrote: >>> -Original Message- >>> From: Lei Wen [mailto:adrian.w...@gmail.com] >>> Sent: Tuesday, March 22, 2011 6:14 PM >>> To: Prafulla Wadaskar >>> Cc: Lei Wen; Heiko Schocher; Wolfgang Denk; u-boot@lists.denx.de; Marek >>> Vasut; Ashish Karkare; Prabhanjan Sarnaik; Yu Tang >>> Subject: Re: [PATCH V3 1/5] pxa: move i2c driver to the common place >>> >> ...snip... > drivers/i2c/Makefile | 1 + > drivers/i2c/mv_i2c.c | 452 > +++ > include/configs/innokom.h | 1 + > include/configs/xm250.h | 1 + > 6 files changed, 455 insertions(+), 470 deletions(-) > delete mode 100644 arch/arm/cpu/pxa/i2c.c > create mode 100644 drivers/i2c/mv_i2c.c ...snip... > -#endif /* CONFIG_HARD_I2C */ > diff --git a/drivers/i2c/Makefile b/drivers/i2c/Makefile > index 052fe36..00a12cc 100644 > --- a/drivers/i2c/Makefile > +++ b/drivers/i2c/Makefile > @@ -29,6 +29,7 @@ COBJS-$(CONFIG_BFIN_TWI_I2C) += bfin-twi_i2c.o > COBJS-$(CONFIG_DRIVER_DAVINCI_I2C) += davinci_i2c.o > COBJS-$(CONFIG_FSL_I2C) += fsl_i2c.o > COBJS-$(CONFIG_I2C_MVTWSI) += mvtwsi.o > +COBJS-$(CONFIG_I2C_MV) += mv_i2c.o Mvtwsi and mv_i2c are two i2c drivers for Marvell. Can you merge these two? >>> As I explain to you before. Although kirkwood and pxa series are both >>> the product >>> of Marvell, but it don't necessary means that they must have the same >>> controller >>> for both product line. For the i2c part, they just use two different >>> controller. >>> So why you keep request merge those two? Do you mean you want to >>> create a unique I2C >>> framework for whole i2c drivers in drivers/i2c? >> >> Hi Lei >> >> 1. Most of i2c drivers supported in u-boot are either in SoC specific folder >> or in drivers/i2c folder, there is no as such thumb rule here. > > New drivers should go to drivers/i2c ! > The existing (old) drivers are just not moved ... > patches welcome! > >> 2. Secondly all these drivers have some common code, mostly i2c_read, >> i2c_write, i2c_probe, etc.. >> 3. Specific to Marvell, we already have mvtwsi.c that supports Kirkwood and >> Orion5X SoCs. Whereas you are adding new mvi2c.c that will support >> armada100, pantheon apart from pxa. >> 4. What about if we need to support some new Marvell SoC with different i2C >> controller? Do we add one more driver? >> >> I would love if some one creates drivers/i2c/i2c_core.c??? not necessarily >> you ;-) >> >> Here is what I would like to suggest. >> 1. cmd_i2c mostly interfaced with i2c_probe, i2c_read, i2c_write, >> i2c_get_bun_num, i2c_set_bus_num, those should go in drivers/i2c_core.c > > Yep, but see below comment. > >> 2. APIs like i2c_start, i2c_stop, i2c_send, i2c_recv, i2c_reset are more I2C >> controller specific and those will be different implementation on different >> SoCs, those can go in SoC specific i2c driver file. > > Yep. > >> 3. all I2C driver files should be in drivers/i2c/ > > Yep. > >> 4. i2c_read/write API need to be redefined since those are not generic to be >> used to access any I2C peripheral( most of the device don't need address to >> be programmed) > > With which devices do you have problems? You can set with > "i2c mw chip address.0 ..." an addresslen = 0 ... or? > >> 5. Flags must be provided for i2c_read/write APIs to have precise control to >> execute I2C_START/I2C_STOP sequence in the call. > > If needed, yes. > >> Since you are the one starting with re-using pxa driver for armada100 and >> Pantheon SoC, why don't you split it into i2c_core.c and i2c_pxa.c? then add >> i2c_armada100.c and i2c_pantheon.c? >> Others can migrate in the similar way. (even mvtwsi,c) >> >> Hi Heiko >> What do you think on this? > > I made such a i2c_core.c file in the multibus/multiadapter branch > for the i2c subsystem, see here: > > http://git.denx.de/?p=u-boot/u-boot-i2c.git;a=shortlog;h=refs/heads/multibus_v2 > > (also you can grep in u-boot ML for discussions about) > > Actual state: > - arm boards: i2c driver tested on suen3 (kirkwood based board) > - powerpc boards: i2c driver tested on mpc82xx, 83xx, 8xx boards > - all others just coded not tested ... > (A result of lacking hw and/or time) > > ToDo: > - rebase against current head > (Sorry, didn;t found time to rebase it since Oktober 2010) > - Update README > - porting arrived new i2c drivers, boards since Oktober 2010 > to this new i2c approach > - testing, testing, testing ... Testers welcome! > > I prefer to integrate this to mainline, before we do above steps > (4?) and 5. As Lei mentioned, if a soc/board has different i2c > controllers and more than one bus we *need* this approach, > so it is not worthwhile to introduce a i2c_core file only ... > instead we should forwarding this branch to mainline? > > Patches are welcome ;-) > >
Re: [U-Boot] [PATCH V3 1/5] pxa: move i2c driver to the common place
Hello Prafulla, Prafulla Wadaskar wrote: >> -Original Message- >> From: Lei Wen [mailto:adrian.w...@gmail.com] >> Sent: Tuesday, March 22, 2011 6:14 PM >> To: Prafulla Wadaskar >> Cc: Lei Wen; Heiko Schocher; Wolfgang Denk; u-boot@lists.denx.de; Marek >> Vasut; Ashish Karkare; Prabhanjan Sarnaik; Yu Tang >> Subject: Re: [PATCH V3 1/5] pxa: move i2c driver to the common place >> > ...snip... drivers/i2c/Makefile |1 + drivers/i2c/mv_i2c.c | 452 +++ include/configs/innokom.h |1 + include/configs/xm250.h |1 + 6 files changed, 455 insertions(+), 470 deletions(-) delete mode 100644 arch/arm/cpu/pxa/i2c.c create mode 100644 drivers/i2c/mv_i2c.c >>> >>> ...snip... >>> -#endif /* CONFIG_HARD_I2C */ diff --git a/drivers/i2c/Makefile b/drivers/i2c/Makefile index 052fe36..00a12cc 100644 --- a/drivers/i2c/Makefile +++ b/drivers/i2c/Makefile @@ -29,6 +29,7 @@ COBJS-$(CONFIG_BFIN_TWI_I2C) += bfin-twi_i2c.o COBJS-$(CONFIG_DRIVER_DAVINCI_I2C) += davinci_i2c.o COBJS-$(CONFIG_FSL_I2C) += fsl_i2c.o COBJS-$(CONFIG_I2C_MVTWSI) += mvtwsi.o +COBJS-$(CONFIG_I2C_MV) += mv_i2c.o >>> Mvtwsi and mv_i2c are two i2c drivers for Marvell. >>> Can you merge these two? >> As I explain to you before. Although kirkwood and pxa series are both >> the product >> of Marvell, but it don't necessary means that they must have the same >> controller >> for both product line. For the i2c part, they just use two different >> controller. >> So why you keep request merge those two? Do you mean you want to >> create a unique I2C >> framework for whole i2c drivers in drivers/i2c? > > Hi Lei > > 1. Most of i2c drivers supported in u-boot are either in SoC specific folder > or in drivers/i2c folder, there is no as such thumb rule here. New drivers should go to drivers/i2c ! The existing (old) drivers are just not moved ... patches welcome! > 2. Secondly all these drivers have some common code, mostly i2c_read, > i2c_write, i2c_probe, etc.. > 3. Specific to Marvell, we already have mvtwsi.c that supports Kirkwood and > Orion5X SoCs. Whereas you are adding new mvi2c.c that will support armada100, > pantheon apart from pxa. > 4. What about if we need to support some new Marvell SoC with different i2C > controller? Do we add one more driver? > > I would love if some one creates drivers/i2c/i2c_core.c??? not necessarily > you ;-) > > Here is what I would like to suggest. > 1. cmd_i2c mostly interfaced with i2c_probe, i2c_read, i2c_write, > i2c_get_bun_num, i2c_set_bus_num, those should go in drivers/i2c_core.c Yep, but see below comment. > 2. APIs like i2c_start, i2c_stop, i2c_send, i2c_recv, i2c_reset are more I2C > controller specific and those will be different implementation on different > SoCs, those can go in SoC specific i2c driver file. Yep. > 3. all I2C driver files should be in drivers/i2c/ Yep. > 4. i2c_read/write API need to be redefined since those are not generic to be > used to access any I2C peripheral( most of the device don't need address to > be programmed) With which devices do you have problems? You can set with "i2c mw chip address.0 ..." an addresslen = 0 ... or? > 5. Flags must be provided for i2c_read/write APIs to have precise control to > execute I2C_START/I2C_STOP sequence in the call. If needed, yes. > Since you are the one starting with re-using pxa driver for armada100 and > Pantheon SoC, why don't you split it into i2c_core.c and i2c_pxa.c? then add > i2c_armada100.c and i2c_pantheon.c? > Others can migrate in the similar way. (even mvtwsi,c) > > Hi Heiko > What do you think on this? I made such a i2c_core.c file in the multibus/multiadapter branch for the i2c subsystem, see here: http://git.denx.de/?p=u-boot/u-boot-i2c.git;a=shortlog;h=refs/heads/multibus_v2 (also you can grep in u-boot ML for discussions about) Actual state: - arm boards: i2c driver tested on suen3 (kirkwood based board) - powerpc boards: i2c driver tested on mpc82xx, 83xx, 8xx boards - all others just coded not tested ... (A result of lacking hw and/or time) ToDo: - rebase against current head (Sorry, didn;t found time to rebase it since Oktober 2010) - Update README - porting arrived new i2c drivers, boards since Oktober 2010 to this new i2c approach - testing, testing, testing ... Testers welcome! I prefer to integrate this to mainline, before we do above steps (4?) and 5. As Lei mentioned, if a soc/board has different i2c controllers and more than one bus we *need* this approach, so it is not worthwhile to introduce a i2c_core file only ... instead we should forwarding this branch to mainline? Patches are welcome ;-) I am afraid, this would get such a big cut as the arm relocation changes ... and it affects all archs. bye, Heiko -- DENX Software Engineering GmbH, MD: Wolfgang Denk & Detlev Zundel HRB 165235 Munich, Office: Kirchen
Re: [U-Boot] [PATCH V3 1/5] pxa: move i2c driver to the common place
> -Original Message- > From: Lei Wen [mailto:adrian.w...@gmail.com] > Sent: Tuesday, March 22, 2011 6:14 PM > To: Prafulla Wadaskar > Cc: Lei Wen; Heiko Schocher; Wolfgang Denk; u-boot@lists.denx.de; Marek > Vasut; Ashish Karkare; Prabhanjan Sarnaik; Yu Tang > Subject: Re: [PATCH V3 1/5] pxa: move i2c driver to the common place > ...snip... > >> drivers/i2c/Makefile | 1 + > >> drivers/i2c/mv_i2c.c | 452 > >> +++ > >> include/configs/innokom.h | 1 + > >> include/configs/xm250.h | 1 + > >> 6 files changed, 455 insertions(+), 470 deletions(-) > >> delete mode 100644 arch/arm/cpu/pxa/i2c.c > >> create mode 100644 drivers/i2c/mv_i2c.c > > > > > > ...snip... > > > >> -#endif /* CONFIG_HARD_I2C */ > >> diff --git a/drivers/i2c/Makefile b/drivers/i2c/Makefile > >> index 052fe36..00a12cc 100644 > >> --- a/drivers/i2c/Makefile > >> +++ b/drivers/i2c/Makefile > >> @@ -29,6 +29,7 @@ COBJS-$(CONFIG_BFIN_TWI_I2C) += bfin-twi_i2c.o > >> COBJS-$(CONFIG_DRIVER_DAVINCI_I2C) += davinci_i2c.o > >> COBJS-$(CONFIG_FSL_I2C) += fsl_i2c.o > >> COBJS-$(CONFIG_I2C_MVTWSI) += mvtwsi.o > >> +COBJS-$(CONFIG_I2C_MV) += mv_i2c.o > > > > Mvtwsi and mv_i2c are two i2c drivers for Marvell. > > Can you merge these two? > > As I explain to you before. Although kirkwood and pxa series are both > the product > of Marvell, but it don't necessary means that they must have the same > controller > for both product line. For the i2c part, they just use two different > controller. > So why you keep request merge those two? Do you mean you want to > create a unique I2C > framework for whole i2c drivers in drivers/i2c? Hi Lei 1. Most of i2c drivers supported in u-boot are either in SoC specific folder or in drivers/i2c folder, there is no as such thumb rule here. 2. Secondly all these drivers have some common code, mostly i2c_read, i2c_write, i2c_probe, etc.. 3. Specific to Marvell, we already have mvtwsi.c that supports Kirkwood and Orion5X SoCs. Whereas you are adding new mvi2c.c that will support armada100, pantheon apart from pxa. 4. What about if we need to support some new Marvell SoC with different i2C controller? Do we add one more driver? I would love if some one creates drivers/i2c/i2c_core.c??? not necessarily you ;-) Here is what I would like to suggest. 1. cmd_i2c mostly interfaced with i2c_probe, i2c_read, i2c_write, i2c_get_bun_num, i2c_set_bus_num, those should go in drivers/i2c_core.c 2. APIs like i2c_start, i2c_stop, i2c_send, i2c_recv, i2c_reset are more I2C controller specific and those will be different implementation on different SoCs, those can go in SoC specific i2c driver file. 3. all I2C driver files should be in drivers/i2c/ 4. i2c_read/write API need to be redefined since those are not generic to be used to access any I2C peripheral( most of the device don't need address to be programmed) 5. Flags must be provided for i2c_read/write APIs to have precise control to execute I2C_START/I2C_STOP sequence in the call. Since you are the one starting with re-using pxa driver for armada100 and Pantheon SoC, why don't you split it into i2c_core.c and i2c_pxa.c? then add i2c_armada100.c and i2c_pantheon.c? Others can migrate in the similar way. (even mvtwsi,c) Hi Heiko What do you think on this? Regards.. Prafulla . . ___ U-Boot mailing list U-Boot@lists.denx.de http://lists.denx.de/mailman/listinfo/u-boot
Re: [U-Boot] [PATCH V3 1/5] pxa: move i2c driver to the common place
Hi Prafulla, On Tue, Mar 22, 2011 at 7:42 PM, Prafulla Wadaskar wrote: > > >> -Original Message- >> From: Lei Wen [mailto:lei...@marvell.com] >> Sent: Thursday, March 17, 2011 12:15 PM >> To: Heiko Schocher; Wolfgang Denk; Prafulla Wadaskar; u- >> b...@lists.denx.de; Marek Vasut; Ashish Karkare; Prabhanjan Sarnaik; >> adrian.w...@gmail.com >> Subject: [PATCH V3 1/5] pxa: move i2c driver to the common place >> >> For better sharing with other platform other than pxa's, >> it is more convenient to put the driver to the common place. >> >> Signed-off-by: Lei Wen >> --- >> Changelog: >> v2: rename previous pxa_i2c to mvi2c. >> >> V3: change previous name from pxa_i2c to mv_i2c >> clean code style issue exist in original code >> >> arch/arm/cpu/pxa/Makefile | 1 - >> arch/arm/cpu/pxa/i2c.c | 469 -- >> --- >> drivers/i2c/Makefile | 1 + >> drivers/i2c/mv_i2c.c | 452 >> +++ >> include/configs/innokom.h | 1 + >> include/configs/xm250.h | 1 + >> 6 files changed, 455 insertions(+), 470 deletions(-) >> delete mode 100644 arch/arm/cpu/pxa/i2c.c >> create mode 100644 drivers/i2c/mv_i2c.c > > > ...snip... > >> -#endif /* CONFIG_HARD_I2C */ >> diff --git a/drivers/i2c/Makefile b/drivers/i2c/Makefile >> index 052fe36..00a12cc 100644 >> --- a/drivers/i2c/Makefile >> +++ b/drivers/i2c/Makefile >> @@ -29,6 +29,7 @@ COBJS-$(CONFIG_BFIN_TWI_I2C) += bfin-twi_i2c.o >> COBJS-$(CONFIG_DRIVER_DAVINCI_I2C) += davinci_i2c.o >> COBJS-$(CONFIG_FSL_I2C) += fsl_i2c.o >> COBJS-$(CONFIG_I2C_MVTWSI) += mvtwsi.o >> +COBJS-$(CONFIG_I2C_MV) += mv_i2c.o > > Mvtwsi and mv_i2c are two i2c drivers for Marvell. > Can you merge these two? As I explain to you before. Although kirkwood and pxa series are both the product of Marvell, but it don't necessary means that they must have the same controller for both product line. For the i2c part, they just use two different controller. So why you keep request merge those two? Do you mean you want to create a unique I2C framework for whole i2c drivers in drivers/i2c? Best regards, Lei ___ U-Boot mailing list U-Boot@lists.denx.de http://lists.denx.de/mailman/listinfo/u-boot
Re: [U-Boot] [PATCH V3 1/5] pxa: move i2c driver to the common place
> -Original Message- > From: Lei Wen [mailto:lei...@marvell.com] > Sent: Thursday, March 17, 2011 12:15 PM > To: Heiko Schocher; Wolfgang Denk; Prafulla Wadaskar; u- > b...@lists.denx.de; Marek Vasut; Ashish Karkare; Prabhanjan Sarnaik; > adrian.w...@gmail.com > Subject: [PATCH V3 1/5] pxa: move i2c driver to the common place > > For better sharing with other platform other than pxa's, > it is more convenient to put the driver to the common place. > > Signed-off-by: Lei Wen > --- > Changelog: > v2: rename previous pxa_i2c to mvi2c. > > V3: change previous name from pxa_i2c to mv_i2c > clean code style issue exist in original code > > arch/arm/cpu/pxa/Makefile |1 - > arch/arm/cpu/pxa/i2c.c| 469 -- > --- > drivers/i2c/Makefile |1 + > drivers/i2c/mv_i2c.c | 452 > +++ > include/configs/innokom.h |1 + > include/configs/xm250.h |1 + > 6 files changed, 455 insertions(+), 470 deletions(-) > delete mode 100644 arch/arm/cpu/pxa/i2c.c > create mode 100644 drivers/i2c/mv_i2c.c ...snip... > -#endif /* CONFIG_HARD_I2C */ > diff --git a/drivers/i2c/Makefile b/drivers/i2c/Makefile > index 052fe36..00a12cc 100644 > --- a/drivers/i2c/Makefile > +++ b/drivers/i2c/Makefile > @@ -29,6 +29,7 @@ COBJS-$(CONFIG_BFIN_TWI_I2C) += bfin-twi_i2c.o > COBJS-$(CONFIG_DRIVER_DAVINCI_I2C) += davinci_i2c.o > COBJS-$(CONFIG_FSL_I2C) += fsl_i2c.o > COBJS-$(CONFIG_I2C_MVTWSI) += mvtwsi.o > +COBJS-$(CONFIG_I2C_MV) += mv_i2c.o Mvtwsi and mv_i2c are two i2c drivers for Marvell. Can you merge these two? Regards.. Prafulla . . ___ U-Boot mailing list U-Boot@lists.denx.de http://lists.denx.de/mailman/listinfo/u-boot
[U-Boot] [PATCH V3 1/5] pxa: move i2c driver to the common place
For better sharing with other platform other than pxa's,
it is more convenient to put the driver to the common place.
Signed-off-by: Lei Wen
---
Changelog:
v2: rename previous pxa_i2c to mvi2c.
V3: change previous name from pxa_i2c to mv_i2c
clean code style issue exist in original code
arch/arm/cpu/pxa/Makefile |1 -
arch/arm/cpu/pxa/i2c.c| 469 -
drivers/i2c/Makefile |1 +
drivers/i2c/mv_i2c.c | 452 +++
include/configs/innokom.h |1 +
include/configs/xm250.h |1 +
6 files changed, 455 insertions(+), 470 deletions(-)
delete mode 100644 arch/arm/cpu/pxa/i2c.c
create mode 100644 drivers/i2c/mv_i2c.c
diff --git a/arch/arm/cpu/pxa/Makefile b/arch/arm/cpu/pxa/Makefile
index 49a6ed3..e8b59a3 100644
--- a/arch/arm/cpu/pxa/Makefile
+++ b/arch/arm/cpu/pxa/Makefile
@@ -28,7 +28,6 @@ LIB = $(obj)lib$(CPU).o
START = start.o
COBJS += cpu.o
-COBJS += i2c.o
COBJS += pxafb.o
COBJS += timer.o
COBJS += usb.o
diff --git a/arch/arm/cpu/pxa/i2c.c b/arch/arm/cpu/pxa/i2c.c
deleted file mode 100644
index 7aa49ae..000
--- a/arch/arm/cpu/pxa/i2c.c
+++ /dev/null
@@ -1,469 +0,0 @@
-/*
- * (C) Copyright 2000
- * Paolo Scaffardi, AIRVENT SAM s.p.a - RIMINI(ITALY), arse...@tin.it
- *
- * (C) Copyright 2000 Sysgo Real-Time Solutions, GmbH
- * Marius Groeger
- *
- * (C) Copyright 2003 Pengutronix e.K.
- * Robert Schwebel
- *
- * See file CREDITS for list of people who contributed to this
- * project.
- *
- * 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; either version 2 of
- * the License, or (at your option) any later version.
- *
- * 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., 59 Temple Place, Suite 330, Boston,
- * MA 02111-1307 USA
- *
- * Back ported to the 8xx platform (from the 8260 platform) by
- * murray.jen...@cmst.csiro.au, 27-Jan-01.
- */
-
-/* FIXME: this file is PXA255 specific! What about other XScales? */
-
-#include
-#include
-
-#ifdef CONFIG_HARD_I2C
-
-/*
- * - CONFIG_SYS_I2C_SPEED
- * - I2C_PXA_SLAVE_ADDR
- */
-
-#include
-#include
-#include
-
-/*#define DEBUG_I2C 1 /###* activate local debugging output
*/
-#define I2C_PXA_SLAVE_ADDR 0x1 /* slave pxa unit address */
-
-#if (CONFIG_SYS_I2C_SPEED == 40)
-#define I2C_ICR_INIT (ICR_FM | ICR_BEIE | ICR_IRFIE | ICR_ITEIE | ICR_GCD |
ICR_SCLE)
-#else
-#define I2C_ICR_INIT (ICR_BEIE | ICR_IRFIE | ICR_ITEIE | ICR_GCD | ICR_SCLE)
-#endif
-
-#define I2C_ISR_INIT 0x7FF
-
-#ifdef DEBUG_I2C
-#define PRINTD(x) printf x
-#else
-#define PRINTD(x)
-#endif
-
-
-/* Shall the current transfer have a start/stop condition? */
-#define I2C_COND_NORMAL0
-#define I2C_COND_START 1
-#define I2C_COND_STOP 2
-
-/* Shall the current transfer be ack/nacked or being waited for it? */
-#define I2C_ACKNAK_WAITACK 1
-#define I2C_ACKNAK_SENDACK 2
-#define I2C_ACKNAK_SENDNAK 4
-
-/* Specify who shall transfer the data (master or slave) */
-#define I2C_READ 0
-#define I2C_WRITE 1
-
-/* All transfers are described by this data structure */
-struct i2c_msg {
- u8 condition;
- u8 acknack;
- u8 direction;
- u8 data;
-};
-
-
-/**
- * i2c_pxa_reset: - reset the host controller
- *
- */
-
-static void i2c_reset( void )
-{
- writel(readl(ICR) & ~ICR_IUE, ICR); /* disable unit */
- writel(readl(ICR) | ICR_UR, ICR); /* reset the unit */
- udelay(100);
- writel(readl(ICR) & ~ICR_IUE, ICR); /* disable unit */
-#ifdef CONFIG_CPU_MONAHANS
- /* | CKENB_1_PWM1 | CKENB_0_PWM0); */
- writel(readl(CKENB) | (CKENB_4_I2C), CKENB);
-#else /* CONFIG_CPU_MONAHANS */
- /* set the global I2C clock on */
- writel(readl(CKEN) | CKEN14_I2C, CKEN);
-#endif
- writel(I2C_PXA_SLAVE_ADDR, ISAR); /* set our slave address */
- writel(I2C_ICR_INIT, ICR); /* set control reg values */
- writel(I2C_ISR_INIT, ISR); /* set clear interrupt bits */
- writel(readl(ICR) | ICR_IUE, ICR); /* enable unit */
- udelay(100);
-}
-
-
-/**
- * i2c_isr_set_cleared: - wait until certain bits of the I2C status register
- * are set and cleared
- *
- * @return: 1 in case of success, 0 means timeout (no match within 10 ms).
- */
-static int i2c_isr_set_cleared( unsigned long set_mask, unsigned long
cleared_mask )
-{
-
