[PATCH v7 RESEND 2/3] thermal: exynos_tmu: Support new hardware and update TMU interface

[손신]

Hello, Tudor Ambarus

> -----Original Message-----
> From: Tudor Ambarus [mailto:tudor.ambarus at linaro.org]
> Sent: Wednesday, November 26, 2025 6:22 PM
> To: 손신 <shin.son at samsung.com>; 'Bartlomiej Zolnierkiewicz'
> <bzolnier at gmail.com>; 'Krzysztof Kozlowski' <krzk at kernel.org>; 'Rafael J .
> Wysocki' <rafael at kernel.org>; 'Daniel Lezcano' <daniel.lezcano at linaro.org>;
> 'Zhang Rui' <rui.zhang at intel.com>; 'Lukasz Luba' <lukasz.luba at arm.com>;
> 'Rob Herring' <robh at kernel.org>; 'Conor Dooley' <conor+dt at kernel.org>;
> 'Alim Akhtar' <alim.akhtar at samsung.com>; youngmin.nam at samsung.com
> Cc: 'Henrik Grimler' <henrik at grimler.se>; linux-pm at vger.kernel.org; linux-
> samsung-soc at vger.kernel.org; devicetree at vger.kernel.org; linux-arm-
> kernel at lists.infradead.org; linux-kernel at vger.kernel.org; 'Peter Griffin'
> <peter.griffin at linaro.org>; 'André Draszik' <andre.draszik at linaro.org>;
> 'William McVicker' <willmcvicker at google.com>; jyescas at google.com
> Subject: Re: [PATCH v7 RESEND 2/3] thermal: exynos_tmu: Support new
> hardware and update TMU interface
> 
> Hi, Shin Son,
> 
> On 11/26/25 9:19 AM, 손신 wrote:
> >> Looking at the exynosautov9 registers that you described and
> >> comparing them with
> >> gs101 I see just 2 differences:
> >> 1/ exnosautov2 has a TRIMINFO_CONFIG2 register, while gs101 doesn't
> >> 2/ EXYNOSAUTOV920_PEND register fields differ from GS101
> > TRIMINFO_CONFIG2 doesn't exist on eav920 either; I simply misnamed it.
> > However, the PEND register indeed differs from GS101.
> >
> >> Given the similarities, and considering the EXYNOS9_ registers rename
> from:
> >> https://lore.kernel.org/linux-samsung-soc/20251117074140.4090939-5-
> >> youngmin.nam at samsung.com/
> >> would it make sense to use the SoC-era name instead of specific SoC,
> i.e.
> >> s/EXYNOSAUTOV920_/EXYNOS9_ and use the latter for both exynosautov9
> >> and gs101?
> >>
> > First of all, as far as I know, EXYNOS9 is not the same as exynosautov9,
> and exynosautov920 also differs from exynosautov9.
> 
> See also see this patch, or maybe the entire patch set:
> https://lore.kernel.org/linux-samsung-soc/20251117074140.4090939-2-
> youngmin.nam at samsung.com/
> 
> It's not just autov9 and gs101 that have similar TMU registers (with the
> two exceptions AFAICT), it's also exynos850 that seems identical with
> autov9.

Yes, Do you have any plans to upstream the GS101 TMU code? From what I understand,
Autov9 and exynos850 are unlikely to be upstreamed in their current form.

> All seem to be part of the same "Exynos9-era" SoCs. Let's think about how
> gs101/exynos850 TMU addition will follow. Shall one use the EXYNOSAUTOV920
> registers? That seems misleading. Shall one redefine the entire register
> set?
> That won't fly because of the code duplication.

I kind of admit that.

> Thus I propose to use the EXYNOS9 prefix for the register definitions, and
> if there are SoCs with slight differences, that can be handled with
> compatible match data and specific SoC definitions, but only where things
> differ.

However, I am not sure whether Exynos2200, 7885, 990, 9810, 8890, 8895, or FSD share the same TMU hardware layout as exynosautov920.
So I’m wondering whether the EXYNOS9 prefix should be limited to GS101 and eav920, or if we should consider a different prefix that better reflects the grouping.

> > So while sharing a common prefix is a good suggestion in general, I
> > believe it's not appropriate here Because the register definitions are
> not fully compatible across these SoCs. Using a common name array may
> introduce confusion later.
> 
> Please reconsider this. Maybe Youngmin Nam or others can intervene.

Ok, I'll reconsider this based on your clarification. Thank you for the detailed feedback.

> 
> Thanks!
> ta

Best regards,
Shin