| Message ID | 20210627220544.8757-1-rdunlap@infradead.org (mailing list archive) |
|---|---|
| Headers | show |
| Series | sh: fixes for various build and kconfig warnings | expand |
Hi Randy! On 6/28/21 12:05 AM, Randy Dunlap wrote: > Fix a few build warnings and one kconfig warning on SUPERH. > > v2: add a previously-sent Kconfig patch > add a few more Cc's > > Cc: Yoshinori Sato <ysato@users.sourceforge.jp> > Cc: Rich Felker <dalias@libc.org> > Cc: linux-sh@vger.kernel.org > Cc: John Paul Adrian Glaubitz <glaubitz@physik.fu-berlin.de> > Cc: Geert Uytterhoeven <geert+renesas@glider.be> > > [PATCH 1/3 v2] sh: fix kconfig unmet dependency warning for FRAME_POINTER > [PATCH 2/3 v2] sh: define __BIG_ENDIAN for math-emu > [PATCH 3/3 v2] sh: fix READ/WRITE redefinition warnings > > arch/sh/Kconfig.debug | 1 > arch/sh/include/asm/sfp-machine.h | 8 +++++ > arch/sh/math-emu/math.c | 44 ++++++++++++++-------------- > 3 files changed, 31 insertions(+), 22 deletions(-) I'll test these tomorrow on my SH-7785LCR board. Would it be possible to queue them up for linux-next after verification? Adrian
On 6/27/21 3:26 PM, John Paul Adrian Glaubitz wrote: > Hi Randy! > > On 6/28/21 12:05 AM, Randy Dunlap wrote: >> Fix a few build warnings and one kconfig warning on SUPERH. >> >> v2: add a previously-sent Kconfig patch >> add a few more Cc's >> >> Cc: Yoshinori Sato <ysato@users.sourceforge.jp> >> Cc: Rich Felker <dalias@libc.org> >> Cc: linux-sh@vger.kernel.org >> Cc: John Paul Adrian Glaubitz <glaubitz@physik.fu-berlin.de> >> Cc: Geert Uytterhoeven <geert+renesas@glider.be> >> >> [PATCH 1/3 v2] sh: fix kconfig unmet dependency warning for FRAME_POINTER >> [PATCH 2/3 v2] sh: define __BIG_ENDIAN for math-emu >> [PATCH 3/3 v2] sh: fix READ/WRITE redefinition warnings >> >> arch/sh/Kconfig.debug | 1 >> arch/sh/include/asm/sfp-machine.h | 8 +++++ >> arch/sh/math-emu/math.c | 44 ++++++++++++++-------------- >> 3 files changed, 31 insertions(+), 22 deletions(-) > > I'll test these tomorrow on my SH-7785LCR board. Would it be possible to queue > them up for linux-next after verification? Thanks. Hopefully they will be queued and put into linux-next, but that's up to the arch/sh/ maintainers, and I haven't heard from them lately. :(
Hi Randy! On 6/28/21 12:30 AM, Randy Dunlap wrote: >>> [PATCH 1/3 v2] sh: fix kconfig unmet dependency warning for FRAME_POINTER >>> [PATCH 2/3 v2] sh: define __BIG_ENDIAN for math-emu >>> [PATCH 3/3 v2] sh: fix READ/WRITE redefinition warnings >>> >>> arch/sh/Kconfig.debug | 1 >>> arch/sh/include/asm/sfp-machine.h | 8 +++++ >>> arch/sh/math-emu/math.c | 44 ++++++++++++++-------------- >>> 3 files changed, 31 insertions(+), 22 deletions(-) >> >> I'll test these tomorrow on my SH-7785LCR board. Would it be possible to queue >> them up for linux-next after verification? > > Thanks. Hopefully they will be queued and put into linux-next, > but that's up to the arch/sh/ maintainers, and I haven't heard > from them lately. :( Whoops, I was meant to direct the second part to Rich :-). Rich is still around and he reviews and merges patches. Sometimes he is a bit busy with other things, so response times can sometimes be a bit delayed. Adrian
Hello! On 6/28/21 12:26 AM, John Paul Adrian Glaubitz wrote: >> arch/sh/Kconfig.debug | 1 >> arch/sh/include/asm/sfp-machine.h | 8 +++++ >> arch/sh/math-emu/math.c | 44 ++++++++++++++-------------- >> 3 files changed, 31 insertions(+), 22 deletions(-) > > I'll test these tomorrow on my SH-7785LCR board. Would it be possible to queue > them up for linux-next after verification? Just as a heads-up: My fast EPYC server for kernel cross-compilation is currently down, so I'll have to set up a new machine at SUSE tomorrow for that purpose. Hope to be able to test the kernel after that. Sorry for the delay, the other server is currently down due to an A/C failure at the university. Adrian
On 6/27/21 3:05 PM, Randy Dunlap wrote: > Fix a few build warnings and one kconfig warning on SUPERH. > > v2: add a previously-sent Kconfig patch > add a few more Cc's > > Cc: Yoshinori Sato <ysato@users.sourceforge.jp> > Cc: Rich Felker <dalias@libc.org> > Cc: linux-sh@vger.kernel.org > Cc: John Paul Adrian Glaubitz <glaubitz@physik.fu-berlin.de> > Cc: Geert Uytterhoeven <geert+renesas@glider.be> > > [PATCH 1/3 v2] sh: fix kconfig unmet dependency warning for FRAME_POINTER > [PATCH 2/3 v2] sh: define __BIG_ENDIAN for math-emu > [PATCH 3/3 v2] sh: fix READ/WRITE redefinition warnings > > arch/sh/Kconfig.debug | 1 > arch/sh/include/asm/sfp-machine.h | 8 +++++ > arch/sh/math-emu/math.c | 44 ++++++++++++++-------------- > 3 files changed, 31 insertions(+), 22 deletions(-) > Hi Rich and Yoshinori, What is the status of arch/sh/ in general and of these patches in particular? thanks.
Hi Randy! On 9/8/21 22:19, Randy Dunlap wrote: > What is the status of arch/sh/ in general and > of these patches in particular? I've also been trying to reach out to Yoshinori and Rich. I know that Yoshinori is currently busy with other work but he can be reached over Twitter [1]. I don't know about Rich though. There are quite a number of patches on the mailing list that need reviewing and I fear if that doesn't happen in the foreseeable future, the SH port is being kicked out which would be a pity given that we're still maintaining the port in Debian and given that there is new hardware available with the J-Core board [2]. Adrian > [1] https://twitter.com/ysat0/ > [2] https://www.cnx-software.com/2017/03/13/turtle-board-is-a-raspberry-pi-2-like-fpga-board-for-j-core-j2-open-source-superh-sh2-soc/
HI Adrian, On Thu, 9 Sept 2021 at 17:25, John Paul Adrian Glaubitz <glaubitz@physik.fu-berlin.de> wrote: > There are quite a number of patches on the mailing list that need reviewing and > I fear if that doesn't happen in the foreseeable future, the SH port is being > kicked out which would be a pity given that we're still maintaining the port in > Debian and given that there is new hardware available with the J-Core board [2]. This really is a poor situation. The fact that there are patches means that it's not totally dead but no one ever looking at them really puts people off bothering in the future. I might have a go at getting OF to work on the SH4 hardware I have but knowing that the patches will probably never get looked at is very demotivating. Can we get a new maintainer from somewhere? Cheers, Daniel
On 9/9/21 3:25 AM, John Paul Adrian Glaubitz wrote: > Hi Randy! > > On 9/8/21 22:19, Randy Dunlap wrote: >> What is the status of arch/sh/ in general and >> of these patches in particular? > > I've also been trying to reach out to Yoshinori and Rich. I know that Yoshinori is > currently busy with other work but he can be reached over Twitter [1]. I don't > know about Rich though. https://twitter.com/richfelker (That said I believe he's on a road trip with his family this week?) > There are quite a number of patches on the mailing list that need reviewing and > I fear if that doesn't happen in the foreseeable future, the SH port is being > kicked out which would be a pity given that we're still maintaining the port in > Debian and given that there is new hardware available with the J-Core board [2]. Rich tends to miss things that go by on the list, or silently assume things will go in through somebody else's tree. That said he responds to email and if all else fails I have his cell phone number, so poke _me_ about it. :) What I _don't_ have is the technical expertise to say more than "yup it compiled and ran on qemu and/or my turtle board". (I have two other pieces of sh4 hardware and an original sh2 board, but they're in austin and I'm in chicago at the moment, and none of them has a particularly friendly bootloader for feeding new kernel images into.) If you just want "it worked", I can sign off on that. If you want "is it a good idea", that's a higher bar. Rob
On Thu, Sep 09, 2021 at 06:08:58PM +0900, Daniel Palmer wrote: > HI Adrian, > > On Thu, 9 Sept 2021 at 17:25, John Paul Adrian Glaubitz > <glaubitz@physik.fu-berlin.de> wrote: > > There are quite a number of patches on the mailing list that need reviewing and > > I fear if that doesn't happen in the foreseeable future, the SH port is being > > kicked out which would be a pity given that we're still maintaining the port in > > Debian and given that there is new hardware available with the J-Core board [2]. > > This really is a poor situation. The fact that there are patches means > that it's not totally dead but no one ever looking at them really puts > people off bothering in the future. > I might have a go at getting OF to work on the SH4 hardware I have but > knowing that the patches will probably never get looked at is very > demotivating. > > Can we get a new maintainer from somewhere? Hi. I see there's a situation that needs my attention here. I will plan to review and merge anything important/blocking that doesn't have problems this week. In the bigger picture, the past few weeks and even months I've been in a sort of "avoid burnout safety mode". :-) Probably partly on account of this pandemic still being a thing because people insist on being stupid. I'm not gone and won't be, but some things that haven't seemed as urgent, including kernel stuff and especially piles of email of mixed importance levels, have gotten pushed back to reduce stress. Please don't hesitate to wave a "hey this is important, come take a quick look!" flag at me if needed. At the same time, I am open to the possibility of a new maintainer or co-maintainer if that ends up being what makes sense. Are there any candidates? Rich
Hi Rich! On 9/12/21 03:57, Rich Felker wrote: > Hi. I see there's a situation that needs my attention here. I will > plan to review and merge anything important/blocking that doesn't have > problems this week. I'm glad to here that you're still active. I will try to help assembling the list of patches. I won't be able to test them though as I'm not at home so I'm unable to reset the machine in case it crashes due to a bad kernel patch. So, basically, I just have one shot free. > In the bigger picture, the past few weeks and even months I've been in > a sort of "avoid burnout safety mode". :-) Probably partly on account > of this pandemic still being a thing because people insist on being > stupid. I'm not gone and won't be, but some things that haven't seemed > as urgent, including kernel stuff and especially piles of email of > mixed importance levels, have gotten pushed back to reduce stress. > Please don't hesitate to wave a "hey this is important, come take a > quick look!" flag at me if needed. We definitely need to get all those patches merged that fix warnings as Linus recently moved to building with -Werror by default. > At the same time, I am open to the possibility of a new maintainer or > co-maintainer if that ends up being what makes sense. Are there any > candidates? I would generally be interested to help although I'm not as knowledgeable when it comes to low-level kernel development. Adrian
On 9/12/21 5:36 AM, John Paul Adrian Glaubitz wrote: > Hi Rich! > > On 9/12/21 03:57, Rich Felker wrote: >> Hi. I see there's a situation that needs my attention here. I will >> plan to review and merge anything important/blocking that doesn't have >> problems this week. > > I'm glad to here that you're still active. I will try to help assembling > the list of patches. I won't be able to test them though as I'm not at > home so I'm unable to reset the machine in case it crashes due to a bad > kernel patch. So, basically, I just have one shot free. I smoketest on qemu-system-sh4 first just because it's so convenient, then test on my j-core (sh2+) turtle board because it's the next most convenient thing (it's USB powered and one lives in my backpack). >> At the same time, I am open to the possibility of a new maintainer or >> co-maintainer if that ends up being what makes sense. Are there any >> candidates? > > I would generally be interested to help although I'm not as knowledgeable > when it comes to low-level kernel development. I'm in a similar position: I can put more cycles into this (and I've collected patch stacks for Rich before), but I don't have the domain expertise to make judgement calls about device tree representation, interrupt routing, nontrivial lock ordering, knowing when you need a memory barrier and why... (Also, I've been running up to 3 days behind on my email recently, and catching up on weekends...) Rob
Hi Rich, On Sun, 12 Sept 2021 at 10:57, Rich Felker <dalias@libc.org> wrote: > > On Thu, Sep 09, 2021 at 06:08:58PM +0900, Daniel Palmer wrote: > Hi. I see there's a situation that needs my attention here. I will > plan to review and merge anything important/blocking that doesn't have > problems this week. Great. If you want to have it tested on real hardware please let me know. I have an ecovec24 (SH4A) board connected to CI so I can easily build and boot test. > At the same time, I am open to the possibility of a new maintainer or > co-maintainer if that ends up being what makes sense. Are there any > candidates? I think the problem is that while there are hobbyists out there the number of people that are still familiar with how the SuperH stuff works is pretty low. I don't have the debug dongles etc anymore so if some treewide change breaks everything then I don't know how I'd debug it. Maybe we could add a few people that still have working hardware as reviewers so they see patches and build test them for you? Cheers, Daniel
On Sun, Sep 12, 2021 at 12:36:38PM +0200, John Paul Adrian Glaubitz wrote: > Hi Rich! > > On 9/12/21 03:57, Rich Felker wrote: > > Hi. I see there's a situation that needs my attention here. I will > > plan to review and merge anything important/blocking that doesn't have > > problems this week. > > I'm glad to here that you're still active. I will try to help assembling > the list of patches. I won't be able to test them though as I'm not at > home so I'm unable to reset the machine in case it crashes due to a bad > kernel patch. So, basically, I just have one shot free. I didn't get through that yet, but I have rebased the patches that were pending in for-next onto v5.15-rc1 (no conflicts) and smoke-tested that a sh4 build runs in my qemu environment. linux-next pulled them 27 hours ago and hasn't complained yet either. I started going through the list/patch backlog, but didn't make it nearly as far as I'd like yet. If you have even a vague list of what's important (warnings breaking the build, unapplied changes blocking removal of cruft from other parts of the kernel and making people unhappy with us, etc.) that would be really helpful. I'll follow up again soon. Rich
Hi Rich! On 9/22/21 04:45, Rich Felker wrote: > I didn't get through that yet, but I have rebased the patches that > were pending in for-next onto v5.15-rc1 (no conflicts) and > smoke-tested that a sh4 build runs in my qemu environment. linux-next > pulled them 27 hours ago and hasn't complained yet either. Sounds promising, thanks! > I started going through the list/patch backlog, but didn't make it > nearly as far as I'd like yet. If you have even a vague list of what's > important (warnings breaking the build, unapplied changes blocking > removal of cruft from other parts of the kernel and making people > unhappy with us, etc.) that would be really helpful. I will start testing kernel updates again from next week as I'm currently busy with other open source work I want to get off my table first. > I'll follow up again soon. Thanks, looking forward for more updates. Adrian
Hi Rich! On 9/22/21 04:45, Rich Felker wrote: > I didn't get through that yet, but I have rebased the patches that > were pending in for-next onto v5.15-rc1 (no conflicts) and > smoke-tested that a sh4 build runs in my qemu environment. linux-next > pulled them 27 hours ago and hasn't complained yet either. Do we need another PR against Linus' tree from your side or are they pulled in automatically? > I started going through the list/patch backlog, but didn't make it > nearly as far as I'd like yet. If you have even a vague list of what's > important (warnings breaking the build, unapplied changes blocking > removal of cruft from other parts of the kernel and making people > unhappy with us, etc.) that would be really helpful. > > I'll follow up again soon. Any news on this? Adrian
Fix a few build warnings and one kconfig warning on SUPERH. v2: add a previously-sent Kconfig patch add a few more Cc's Cc: Yoshinori Sato <ysato@users.sourceforge.jp> Cc: Rich Felker <dalias@libc.org> Cc: linux-sh@vger.kernel.org Cc: John Paul Adrian Glaubitz <glaubitz@physik.fu-berlin.de> Cc: Geert Uytterhoeven <geert+renesas@glider.be> [PATCH 1/3 v2] sh: fix kconfig unmet dependency warning for FRAME_POINTER [PATCH 2/3 v2] sh: define __BIG_ENDIAN for math-emu [PATCH 3/3 v2] sh: fix READ/WRITE redefinition warnings arch/sh/Kconfig.debug | 1 arch/sh/include/asm/sfp-machine.h | 8 +++++ arch/sh/math-emu/math.c | 44 ++++++++++++++-------------- 3 files changed, 31 insertions(+), 22 deletions(-)