* [GIT PULL] firmware: arm_ffa: Fix for v7.1
@ 2026-04-07 10:08 Sudeep Holla
2026-04-11 8:49 ` Krzysztof Kozlowski
0 siblings, 1 reply; 5+ messages in thread
From: Sudeep Holla @ 2026-04-07 10:08 UTC (permalink / raw)
To: ARM SoC Team, SoC Team, ALKML; +Cc: Sudeep Holla, Arnd Bergmann
Hi ARM SoC Team,
Please pull ! This is the only fix/update I have at the moment for v7.1
So, I am sending it early as fix but late as an update for v7.1.
Regards,
Sudeep
-->8
The following changes since commit 6de23f81a5e08be8fbf5e8d7e9febc72a5b5f27f:
Linux 7.0-rc1 (2026-02-22 13:18:59 -0800)
are available in the Git repository at:
git://git.kernel.org/pub/scm/linux/kernel/git/sudeep.holla/linux.git tags/ffa-fix-7.1
for you to fetch changes up to 83210251fd70d5f96bcdc8911e15f7411a6b2463:
firmware: arm_ffa: Use the correct buffer size during RXTX_MAP (2026-04-07 10:47:42 +0100)
----------------------------------------------------------------
Arm FF-A fix for v7.1
Use the page aligned backing allocation size when computing the RXTX_MAP
page count. This fixes FF-A RX/TX buffer registration on kernels built
with 16K/64K PAGE_SIZE, where alloc_pages_exact() backs the buffer with a
larger aligned span than the discovered minimum buffer size.
----------------------------------------------------------------
Sebastian Ene (1):
firmware: arm_ffa: Use the correct buffer size during RXTX_MAP
drivers/firmware/arm_ffa/driver.c | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
^ permalink raw reply [flat|nested] 5+ messages in thread* Re: [GIT PULL] firmware: arm_ffa: Fix for v7.1 2026-04-07 10:08 [GIT PULL] firmware: arm_ffa: Fix for v7.1 Sudeep Holla @ 2026-04-11 8:49 ` Krzysztof Kozlowski 2026-04-11 17:35 ` Sudeep Holla 0 siblings, 1 reply; 5+ messages in thread From: Krzysztof Kozlowski @ 2026-04-11 8:49 UTC (permalink / raw) To: Sudeep Holla; +Cc: ARM SoC Team, SoC Team, ALKML, Arnd Bergmann On Tue, Apr 07, 2026 at 11:08:39AM +0100, Sudeep Holla wrote: > Hi ARM SoC Team, > > Please pull ! This is the only fix/update I have at the moment for v7.1 > So, I am sending it early as fix but late as an update for v7.1. > > Regards, > Sudeep > > -->8 > > The following changes since commit 6de23f81a5e08be8fbf5e8d7e9febc72a5b5f27f: > > Linux 7.0-rc1 (2026-02-22 13:18:59 -0800) > > are available in the Git repository at: > > git://git.kernel.org/pub/scm/linux/kernel/git/sudeep.holla/linux.git tags/ffa-fix-7.1 > > for you to fetch changes up to 83210251fd70d5f96bcdc8911e15f7411a6b2463: > > firmware: arm_ffa: Use the correct buffer size during RXTX_MAP (2026-04-07 10:47:42 +0100) > > ---------------------------------------------------------------- > Arm FF-A fix for v7.1 > > Use the page aligned backing allocation size when computing the RXTX_MAP > page count. This fixes FF-A RX/TX buffer registration on kernels built > with 16K/64K PAGE_SIZE, where alloc_pages_exact() backs the buffer with a > larger aligned span than the discovered minimum buffer size. Can we avoid per-driver trees or pulls? You do maintain also ARM SCMI firmware driver, so this could be sent together? I think you also use the same Git tree, right? Thanks, applied Best regards, Krzysztof ^ permalink raw reply [flat|nested] 5+ messages in thread
* Re: [GIT PULL] firmware: arm_ffa: Fix for v7.1 2026-04-11 8:49 ` Krzysztof Kozlowski @ 2026-04-11 17:35 ` Sudeep Holla 2026-04-13 6:23 ` Arnd Bergmann 0 siblings, 1 reply; 5+ messages in thread From: Sudeep Holla @ 2026-04-11 17:35 UTC (permalink / raw) To: Krzysztof Kozlowski Cc: ARM SoC Team, SoC Team, Sudeep Holla, ALKML, Arnd Bergmann On Sat, Apr 11, 2026 at 10:49:50AM +0200, Krzysztof Kozlowski wrote: > On Tue, Apr 07, 2026 at 11:08:39AM +0100, Sudeep Holla wrote: > > Hi ARM SoC Team, > > > > Please pull ! This is the only fix/update I have at the moment for v7.1 > > So, I am sending it early as fix but late as an update for v7.1. > > > > Regards, > > Sudeep > > > > -->8 > > > > The following changes since commit 6de23f81a5e08be8fbf5e8d7e9febc72a5b5f27f: > > > > Linux 7.0-rc1 (2026-02-22 13:18:59 -0800) > > > > are available in the Git repository at: > > > > git://git.kernel.org/pub/scm/linux/kernel/git/sudeep.holla/linux.git tags/ffa-fix-7.1 > > > > for you to fetch changes up to 83210251fd70d5f96bcdc8911e15f7411a6b2463: > > > > firmware: arm_ffa: Use the correct buffer size during RXTX_MAP (2026-04-07 10:47:42 +0100) > > > > ---------------------------------------------------------------- > > Arm FF-A fix for v7.1 > > > > Use the page aligned backing allocation size when computing the RXTX_MAP > > page count. This fixes FF-A RX/TX buffer registration on kernels built > > with 16K/64K PAGE_SIZE, where alloc_pages_exact() backs the buffer with a > > larger aligned span than the discovered minimum buffer size. > > Can we avoid per-driver trees or pulls? You do maintain also ARM SCMI > firmware driver, so this could be sent together? I think you also use > the same Git tree, right? > Sure, I can put all of the firmware drivers I maintain together. I had for some reason assumed individual PR is preferred. -- Regards, Sudeep ^ permalink raw reply [flat|nested] 5+ messages in thread
* Re: [GIT PULL] firmware: arm_ffa: Fix for v7.1 2026-04-11 17:35 ` Sudeep Holla @ 2026-04-13 6:23 ` Arnd Bergmann 2026-04-13 8:32 ` Sudeep Holla 0 siblings, 1 reply; 5+ messages in thread From: Arnd Bergmann @ 2026-04-13 6:23 UTC (permalink / raw) To: Sudeep Holla, Krzysztof Kozlowski; +Cc: arm, SoC Team, ALKML On Sat, Apr 11, 2026, at 19:35, Sudeep Holla wrote: > On Sat, Apr 11, 2026 at 10:49:50AM +0200, Krzysztof Kozlowski wrote: >> On Tue, Apr 07, 2026 at 11:08:39AM +0100, Sudeep Holla wrote: >> > ---------------------------------------------------------------- >> > Arm FF-A fix for v7.1 >> > >> > Use the page aligned backing allocation size when computing the RXTX_MAP >> > page count. This fixes FF-A RX/TX buffer registration on kernels built >> > with 16K/64K PAGE_SIZE, where alloc_pages_exact() backs the buffer with a >> > larger aligned span than the discovered minimum buffer size. >> >> Can we avoid per-driver trees or pulls? You do maintain also ARM SCMI >> firmware driver, so this could be sent together? I think you also use >> the same Git tree, right? > > Sure, I can put all of the firmware drivers I maintain together. I had > for some reason assumed individual PR is preferred. To me, that's a function of how complex the changes are and how you describe them in the changelog text: If you have a lot of changes for the merge window, having one branch per firmware type probably works best, or even multiple ones if you have a series that implements something new and a number of random changes do existing code. If you have only a handful of bugfixes across multiple firmware subsystems, a single 'firmware fixes' is less work for all of us, with no loss of readability in the git history. Arnd ^ permalink raw reply [flat|nested] 5+ messages in thread
* Re: [GIT PULL] firmware: arm_ffa: Fix for v7.1 2026-04-13 6:23 ` Arnd Bergmann @ 2026-04-13 8:32 ` Sudeep Holla 0 siblings, 0 replies; 5+ messages in thread From: Sudeep Holla @ 2026-04-13 8:32 UTC (permalink / raw) To: Arnd Bergmann; +Cc: Krzysztof Kozlowski, arm, SoC Team, Sudeep Holla, ALKML On Mon, Apr 13, 2026 at 08:23:58AM +0200, Arnd Bergmann wrote: > On Sat, Apr 11, 2026, at 19:35, Sudeep Holla wrote: > > On Sat, Apr 11, 2026 at 10:49:50AM +0200, Krzysztof Kozlowski wrote: > >> On Tue, Apr 07, 2026 at 11:08:39AM +0100, Sudeep Holla wrote: > >> > ---------------------------------------------------------------- > >> > Arm FF-A fix for v7.1 > >> > > >> > Use the page aligned backing allocation size when computing the RXTX_MAP > >> > page count. This fixes FF-A RX/TX buffer registration on kernels built > >> > with 16K/64K PAGE_SIZE, where alloc_pages_exact() backs the buffer with a > >> > larger aligned span than the discovered minimum buffer size. > >> > >> Can we avoid per-driver trees or pulls? You do maintain also ARM SCMI > >> firmware driver, so this could be sent together? I think you also use > >> the same Git tree, right? > > > > Sure, I can put all of the firmware drivers I maintain together. I had > > for some reason assumed individual PR is preferred. > > To me, that's a function of how complex the changes are and how > you describe them in the changelog text: If you have a lot of changes > for the merge window, having one branch per firmware type probably > works best, or even multiple ones if you have a series that implements > something new and a number of random changes do existing code. > > If you have only a handful of bugfixes across multiple firmware > subsystems, a single 'firmware fixes' is less work for all of > us, with no loss of readability in the git history. > Understood, will try to follow something along these in the future. -- Regards, Sudeep ^ permalink raw reply [flat|nested] 5+ messages in thread
end of thread, other threads:[~2026-04-13 8:32 UTC | newest] Thread overview: 5+ messages (download: mbox.gz follow: Atom feed -- links below jump to the message on this page -- 2026-04-07 10:08 [GIT PULL] firmware: arm_ffa: Fix for v7.1 Sudeep Holla 2026-04-11 8:49 ` Krzysztof Kozlowski 2026-04-11 17:35 ` Sudeep Holla 2026-04-13 6:23 ` Arnd Bergmann 2026-04-13 8:32 ` Sudeep Holla
This is a public inbox, see mirroring instructions for how to clone and mirror all data and code used for this inbox