Re: [PATCH 6.1 000/568] 6.1.159-rc1 review

[Greg Kroah-Hartman <gregkh@linuxfoundation.org>]

On Thu, Dec 04, 2025 at 05:51:18AM -0800, Ron Economos wrote:
> On 12/4/25 04:06, Mark Brown wrote:
> > On Wed, Dec 03, 2025 at 04:20:02PM +0100, Greg Kroah-Hartman wrote:
> > > This is the start of the stable review cycle for the 6.1.159 release.
> > > There are 568 patches in this series, all will be posted as a response
> > > to this one.  If anyone has any issues with these being applied, please
> > > let me know.
> > I'm seeing a bunch of systems start failing with this release, they
> > start OOMing when previously they ran OK.  Most of them aren't exactly
> > overburned with memory.  These failures bisect to 61717acddadf66
> > (mm/memory: do not populate page table entries beyond i_size), sample
> > bisect from one of the systems including links to test jobs (the bisects
> > for other systems/test sets look very similar):
> > 
> > # bad: [abd89c70c9382759c14c5e7e0b383c2a19594c5c] Linux 6.1.159-rc1
> > # good: [f6e38ae624cf7eb96fb444a8ca2d07caa8d9c8fe] Linux 6.1.158
> > git bisect start 'abd89c70c9382759c14c5e7e0b383c2a19594c5c' 'f6e38ae624cf7eb96fb444a8ca2d07caa8d9c8fe'
> > # test job: [abd89c70c9382759c14c5e7e0b383c2a19594c5c] https://lava.sirena.org.uk/scheduler/job/2168338
> > # bad: [abd89c70c9382759c14c5e7e0b383c2a19594c5c] Linux 6.1.159-rc1
> > git bisect bad abd89c70c9382759c14c5e7e0b383c2a19594c5c
> > # test job: [43c650106e8558fa7cfec5a2e9c8de29233b6552] https://lava.sirena.org.uk/scheduler/job/2168373
> > # good: [43c650106e8558fa7cfec5a2e9c8de29233b6552] rtc: pcf2127: clear minute/second interrupt
> > git bisect good 43c650106e8558fa7cfec5a2e9c8de29233b6552
> > # test job: [b56fbe428919e8c1a548f331d20b8c4608008845] https://lava.sirena.org.uk/scheduler/job/2168393
> > # good: [b56fbe428919e8c1a548f331d20b8c4608008845] net/mlx5e: Preserve shared buffer capacity during headroom updates
> > git bisect good b56fbe428919e8c1a548f331d20b8c4608008845
> > # test job: [445097729a995f87ff7c80d5a161c7e1b5456628] https://lava.sirena.org.uk/scheduler/job/2169640
> > # bad: [445097729a995f87ff7c80d5a161c7e1b5456628] platform/x86: intel: punit_ipc: fix memory corruption
> > git bisect bad 445097729a995f87ff7c80d5a161c7e1b5456628
> > # test job: [ad3b2ce45cce79ddaff01c977d0867d079fa8349] https://lava.sirena.org.uk/scheduler/job/2169710
> > # good: [ad3b2ce45cce79ddaff01c977d0867d079fa8349] kernel.h: Move ARRAY_SIZE() to a separate header
> > git bisect good ad3b2ce45cce79ddaff01c977d0867d079fa8349
> > # test job: [de07228674e9cee27f679ebcf8562f7e3b2cda21] https://lava.sirena.org.uk/scheduler/job/2169731
> > # good: [de07228674e9cee27f679ebcf8562f7e3b2cda21] mptcp: decouple mptcp fastclose from tcp close
> > git bisect good de07228674e9cee27f679ebcf8562f7e3b2cda21
> > # test job: [dca2a95e4ed753ed33da11d7bb78157441d69bad] https://lava.sirena.org.uk/scheduler/job/2169741
> > # good: [dca2a95e4ed753ed33da11d7bb78157441d69bad] pmdomain: samsung: plug potential memleak during probe
> > git bisect good dca2a95e4ed753ed33da11d7bb78157441d69bad
> > # test job: [61717acddadf660fa6969027bfa0d6fd38f8e3e2] https://lava.sirena.org.uk/scheduler/job/2170912
> > # bad: [61717acddadf660fa6969027bfa0d6fd38f8e3e2] mm/memory: do not populate page table entries beyond i_size
> > git bisect bad 61717acddadf660fa6969027bfa0d6fd38f8e3e2
> > # test job: [0de5c14c8e753a547d158530c37efb245f7b40ec] https://lava.sirena.org.uk/scheduler/job/2171171
> > # good: [0de5c14c8e753a547d158530c37efb245f7b40ec] pmdomain: imx: Fix reference count leak in imx_gpc_remove
> > git bisect good 0de5c14c8e753a547d158530c37efb245f7b40ec
> > # test job: [1457e122dd70574a0ca895eea6d6c12ba91312bf] https://lava.sirena.org.uk/scheduler/job/2171268
> > # good: [1457e122dd70574a0ca895eea6d6c12ba91312bf] filemap: cap PTE range to be created to allowed zero fill in folio_map_range()
> > git bisect good 1457e122dd70574a0ca895eea6d6c12ba91312bf
> > # first bad commit: [61717acddadf660fa6969027bfa0d6fd38f8e3e2] mm/memory: do not populate page table entries beyond i_size
> 
> This patch "mm/memory: do not populate page table entries beyond i_size"
> also causes an Oops on RISC-V.
> 
> [    5.940397] Unable to handle kernel paging request at virtual address
> ffffffc6fe000028
> [    5.947616] Oops [#1]
> [    5.949814] Modules linked in:
> [    5.952857] CPU: 0 PID: 147 Comm: exe Not tainted 6.1.158 #2
> [    5.958500] Hardware name: SiFive HiFive Unmatched A00 (DT)
> [    5.964060] epc : _raw_spin_lock+0x12/0x84
> [    5.968141]  ra : filemap_map_pages+0x23e/0x524
> [    5.972658] epc : ffffffff80b4d500 ra : ffffffff801e3d60 sp :
> ffffffd88c923c80
> [    5.979870]  gp : ffffffff81a3f228 tp : ffffffd88c8c1a80 t0 :
> 0000000000000000
> [    5.987078]  t1 : 0000000000000000 t2 : 0000000000000000 s0 :
> ffffffd88c923c90
> [    5.994287]  s1 : 0000000000000000 a0 : ffffffc6fe000028 a1 :
> ffffffc800000000
> [    6.001497]  a2 : 0000000100000000 a3 : 0000000000080000 a4 :
> 0000000000000ff8
> [    6.008706]  a5 : 0000000000010000 a6 : ffffffff813d5858 a7 :
> 0000000000000000
> [    6.015915]  s2 : 0000003f807ff000 s3 : ffffffd88c923d98 s4 :
> 0000000000000000
> [    6.023125]  s5 : ffffffd77fe00ff8 s6 : 0000000000000000 s7 :
> 0000000000000000
> [    6.030333]  s8 : 0000003f80829008 s9 : fffffffffffffffe s10:
> ffffffd880f866f8
> [    6.037543]  s11: ffffffc7020b67c0 t3 : ffffffffffffffff t4 :
> 0000003f807ff000
> [    6.044752]  t5 : 0000000000000000 t6 : 0000000000000000
> [    6.050047] status: 0000000200000120 badaddr: ffffffc6fe000028 cause:
> 000000000000000f
> [    6.057957] [<ffffffff80b4d500>] _raw_spin_lock+0x12/0x84
> [    6.063338] [<ffffffff801e3d60>] filemap_map_pages+0x23e/0x524
> [    6.069157] [<ffffffff80229ffc>] __handle_mm_fault+0xd44/0x1818
> [    6.075064] [<ffffffff8022ab84>] handle_mm_fault+0xb4/0x1b8
> [    6.080622] [<ffffffff8000cd54>] do_page_fault+0x142/0x462
> [    6.086095] [<ffffffff80003f20>] ret_from_exception+0x0/0x16
> [    6.091798] ---[ end trace 0000000000000000 ]---

Thanks all, I'll go drop this now and push out a -rc2 soon.

greg k-h