Date: Wed, 14 Aug 2024 21:00:16 +0200 From: Joshua Rogers <megamansec@gmail.com> To: freebsd-acpi@freebsd.org Subject: Re: suspend/resume on MacbookPro11.4 without X Message-ID: <CA%2B17n5vNPhQUrkA5ndZmj523LbZBBzs=zFVrjdHvACrd0zjoaw@mail.gmail.com> In-Reply-To: <E6D110A1-7058-4F7C-A7BC-9806604237C5@gmail.com> References: <E6D110A1-7058-4F7C-A7BC-9806604237C5@gmail.com>
next in thread | previous in thread | raw e-mail | index | archive | help
--000000000000dba6fe061fa95685 Content-Type: text/plain; charset="UTF-8" Content-Transfer-Encoding: quoted-printable Ah, I see this is specifically mentioned at https://docs.freebsd.org/en/books/handbook/config/index.html#acpi-overview: " In order for the suspend/resume functionality to work correctly the graphics drivers must be loaded on the system." So, I need to ensure that X is loaded before the system is ever suspended (like the lid being closed). Is there any way to ensure that X is loaded before hw.acpi.power_button_state or hw.acpi.lid_switch_state cause a suspension? On Wed, Aug 7, 2024 at 8:53=E2=80=AFPM Joshua Rogers <megamansec@gmail.com>= wrote: > Hi all, > > I'm currently battling with a Macbook Pro 11.4, attempting to get all of > its hardware supported on FreeBSD. At the moment, I'm investigating wheth= er > it is possible to get suspend/resume working at on 13.3-RELEASE-p4. > > I've exhibited a few different behaviors: > > Using the inbuilt SSD/NVMe ("SM951 AHCI"), I exhibit the following > behaviour: > > 1A. No X, Using debug.acpi.suspend_bounce=3D1: > Successfully resumes, but the disk becomes unreadable/unwritable. The > keyboard is responsive but the mouse is not. I can continue some actions, > but clearly only what is cached in memory. I don't know whether this is > expected: I'm using ZFS with full disk encryption. > > 1B. No X, using debug.acpi.suspend_bounce=3D0: > The screen does not turn back on: but everything else seems to work like > beep, drive read/writing, keyboard, etc (I'm basically just typing in the > blind). > > > 2A. Using X and suspend_bounce=3D1: > The screen resumes but the disk becomes unreadable writable. The keyboard > is responsive but the mouse is not. The same as without X > > 2B. Using X and debug.acpi.suspend_bounce=3D0: > Everything works as well as expected. Keyboard and mouse work. Writing to > the disk works. > > --- > > For the non-X use cases, I've built the kernel with the following minimal > configuration [config.txt]. A more minimal version may work, but I didn't > bother to trim it down any further. The behavior exhibited by the > suspension is the same as when built normally. > > At the moment, case 2B is fine: I use X most of the time, and I don't use > suspend_bounce. But I would like to fiz the 1B case at least (no X). Are > there any suggestions of how to debug this? > > FYI: It looks like the Linux kernel went through an issue with this > version of the Macbook too: > https://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git/commit= /?id=3D13cfc732160f > / https://bugzilla.kernel.org/show_bug.cgi?id=3D103211. I've attempted t= o > recreate the patch and it loads, but does not seem to do anything > [asmc-patch.c.patch]. I have no idea if it is relevant or not. > > An initial HW_PROBE can be found here: > https://bsd-hardware.info/?probe=3D6bade1eaf8 > > Cheers, > Joshua. > --000000000000dba6fe061fa95685 Content-Type: text/html; charset="UTF-8" Content-Transfer-Encoding: quoted-printable <div dir=3D"ltr"><div>Ah, I see this is specifically mentioned at <a href= =3D"https://docs.freebsd.org/en/books/handbook/config/index.html#acpi-overv= iew">https://docs.freebsd.org/en/books/handbook/config/index.html#acpi-over= view</a>:</div><div>"=C2=A0=C2=A0 In order for the suspend/resume func= tionality to work correctly the graphics drivers must be loaded on the syst= em."</div><div><br></div><div>So, I need to ensure that X is loaded be= fore the system is ever suspended (like the lid being closed). Is there any= way to ensure that X is loaded before hw.acpi.power_button_state or hw.acp= i.lid_switch_state cause a suspension?<br></div></div><br><div class=3D"gma= il_quote"><div dir=3D"ltr" class=3D"gmail_attr">On Wed, Aug 7, 2024 at 8:53= =E2=80=AFPM Joshua Rogers <<a href=3D"mailto:megamansec@gmail.com">megam= ansec@gmail.com</a>> wrote:<br></div><blockquote class=3D"gmail_quote" s= tyle=3D"margin:0px 0px 0px 0.8ex;border-left:1px solid rgb(204,204,204);pad= ding-left:1ex"><div dir=3D"auto"><span style=3D"background-color:rgb(255,25= 5,255)">Hi all,</span><div dir=3D"ltr"><div><br></div><div>I'm currentl= y battling with a Macbook Pro 11.4, attempting to get all of its hardware s= upported on FreeBSD. At the moment, I'm investigating whether it is pos= sible to get suspend/resume working at on 13.3-RELEASE-p4.</div><div><br></= div><div>I've exhibited a few different behaviors:</div><div><br></div>= <div>Using the inbuilt SSD/NVMe ("SM951 AHCI"), I exhibit the fol= lowing behaviour:</div><div><br></div><div>1A. No X, Using=C2=A0debug.acpi.= suspend_bounce=3D1:</div><div>Successfully resumes, but the disk becomes un= readable/unwritable. The keyboard is responsive but the mouse is not. I can= continue some actions, but clearly only what is cached in memory. I don= 9;t know whether this is expected: I'm using ZFS with full disk encrypt= ion.=C2=A0</div><div><br></div><div>1B. No X, using debug.acpi.suspend_boun= ce=3D0:</div><div>The screen does not turn back on: but everything else see= ms to work like beep, drive read/writing, keyboard, etc (I'm basically = just typing in the blind).=C2=A0</div><div><br></div><div><br></div><div>2A= . Using X and suspend_bounce=3D1:</div><div>The screen resumes but the disk= becomes unreadable writable. The keyboard is responsive but the mouse is n= ot. The same as without X</div><div><br></div><div>2B. Using X and debug.ac= pi.suspend_bounce=3D0:</div><div>Everything works as well as expected. Keyb= oard and mouse work. Writing to the disk works.=C2=A0</div><div><br></div><= div>---</div><div><br></div><div>For the non-X use cases, I've built th= e kernel with the following minimal configuration [config.txt]. A more mini= mal version may work, but I didn't bother to trim it down any further. = The behavior exhibited by the suspension is the same as when built normally= .</div><div><br></div><div>At the moment, case 2B is fine: I use X most of = the time, and I don't use suspend_bounce. But I would like to fiz the 1= B case at least (no X). Are there any suggestions of how to debug this?</di= v><div><br></div><div>FYI: It looks like the Linux kernel went through an i= ssue with this version of the Macbook too:=C2=A0<a href=3D"https://git.kern= el.org/pub/scm/linux/kernel/git/torvalds/linux.git/commit/?id=3D13cfc732160= f" target=3D"_blank">https://git.kernel.org/pub/scm/linux/kernel/git/torval= ds/linux.git/commit/?id=3D13cfc732160f</a>=C2=A0/=C2=A0<a href=3D"https://b= ugzilla.kernel.org/show_bug.cgi?id=3D103211" target=3D"_blank">https://bugz= illa.kernel.org/show_bug.cgi?id=3D103211</a>. I've attempted to recreat= e the patch and it loads, but does not seem to do anything [asmc-patch.c.pa= tch]. I have no idea if it is relevant or not.</div><div><br></div><div>An = initial HW_PROBE can be found here:=C2=A0<a href=3D"https://bsd-hardware.in= fo/?probe=3D6bade1eaf8" target=3D"_blank">https://bsd-hardware.info/?probe= =3D6bade1eaf8</a></div><div><br></div><div>Cheers,</div><div>Joshua.</div><= /div><div dir=3D"ltr"></div></div></blockquote></div> --000000000000dba6fe061fa95685--
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?CA%2B17n5vNPhQUrkA5ndZmj523LbZBBzs=zFVrjdHvACrd0zjoaw>