Message ID | Zfl9Efxe7DwuU_i3@rric.localdomain (mailing list archive) |
---|---|
State | Handled Elsewhere, archived |
Headers | show |
Series | cxl: Fix use of phys_to_target_node() outside of init section | expand |
Robert Richter wrote: > Hi Dan, > > patch below. I have not included it into v2 of the SRAT/CEDT changes > as it is cxl specific and can be applied separately. > > Thanks, > > -Robert > > > On 18.03.24 14:26:41, Dan Williams wrote: > > It should also be the case that cxl_acpi needs this: > > > > diff --git a/drivers/cxl/Kconfig b/drivers/cxl/Kconfig > > index 67998dbd1d46..1bf25185c35b 100644 > > --- a/drivers/cxl/Kconfig > > +++ b/drivers/cxl/Kconfig > > @@ -6,6 +6,7 @@ menuconfig CXL_BUS > > select FW_UPLOAD > > select PCI_DOE > > select FIRMWARE_TABLE > > + select NUMA_KEEP_MEMINFO if NUMA > > help > > CXL is a bus that is electrically compatible with PCI Express, but > > layers three protocols on that signalling (CXL.io, CXL.cache, and > > From be5b495980bae41d879909212db02dac0fba978e Mon Sep 17 00:00:00 2001 Hi Robert, When you send inline patches like this can you remember to include a scissors line? That way tools like "b4 am" automatically know where to trim things. So add a line like the following: -- >8 -- ...see "git mailinfo --help" for details. Also note that if you reply with an updated patch in a series include the "vX NN/MM" suffix, like "Subject: [PATCH v3 2/3] ..." so that b4 am knows to perform a "partial reroll". > From: Robert Richter <rrichter@amd.com> > Date: Tue, 19 Mar 2024 09:28:33 +0100 > Subject: [PATCH] cxl: Fix use of phys_to_target_node() outside of init section > > The CXL driver uses both functions phys_to_target_node() and > memory_add_physaddr_to_nid(). The x86 architecture relies on the > NUMA_KEEP_MEMINFO kernel option to be set. Enable the option for the > driver accordingly. > > Suggested-by: Dan Williams <dan.j.williams@intel.com> > Signed-off-by: Robert Richter <rrichter@amd.com> > --- > drivers/cxl/Kconfig | 1 + > 1 file changed, 1 insertion(+) > > diff --git a/drivers/cxl/Kconfig b/drivers/cxl/Kconfig > index 67998dbd1d46..6140b3529a29 100644 > --- a/drivers/cxl/Kconfig > +++ b/drivers/cxl/Kconfig > @@ -6,6 +6,7 @@ menuconfig CXL_BUS > select FW_UPLOAD > select PCI_DOE > select FIRMWARE_TABLE > + select NUMA_KEEP_MEMINFO if (NUMA && X86) > help > CXL is a bus that is electrically compatible with PCI Express, but > layers three protocols on that signalling (CXL.io, CXL.cache, and > -- > 2.39.2 >
On 19.03.24 17:21:53, Dan Williams wrote: > Robert Richter wrote: > > Hi Dan, > > > > patch below. I have not included it into v2 of the SRAT/CEDT changes > > as it is cxl specific and can be applied separately. > > > > Thanks, > > > > -Robert > > > > > > On 18.03.24 14:26:41, Dan Williams wrote: > > > It should also be the case that cxl_acpi needs this: > > > > > > diff --git a/drivers/cxl/Kconfig b/drivers/cxl/Kconfig > > > index 67998dbd1d46..1bf25185c35b 100644 > > > --- a/drivers/cxl/Kconfig > > > +++ b/drivers/cxl/Kconfig > > > @@ -6,6 +6,7 @@ menuconfig CXL_BUS > > > select FW_UPLOAD > > > select PCI_DOE > > > select FIRMWARE_TABLE > > > + select NUMA_KEEP_MEMINFO if NUMA > > > help > > > CXL is a bus that is electrically compatible with PCI Express, but > > > layers three protocols on that signalling (CXL.io, CXL.cache, and > > > > From be5b495980bae41d879909212db02dac0fba978e Mon Sep 17 00:00:00 2001 > > Hi Robert, > > When you send inline patches like this can you remember to include a > scissors line? That way tools like "b4 am" automatically know where to > trim things. So add a line like the following: > > -- >8 -- > > ...see "git mailinfo --help" for details. Thanks for the inside on your patch processing. Will use that in the future. > > Also note that if you reply with an updated patch in a series include > the "vX NN/MM" suffix, like "Subject: [PATCH v3 2/3] ..." so that b4 am > knows to perform a "partial reroll". This patch is in addition to the other SRAT patches and can be applied directly to the cxl tree. That is why there is no version update here. But I replied to this series for reference. I saw the b4 shazam --no-parent option, would that help here? Thanks, -Robert > > > From: Robert Richter <rrichter@amd.com> > > Date: Tue, 19 Mar 2024 09:28:33 +0100 > > Subject: [PATCH] cxl: Fix use of phys_to_target_node() outside of init section > > > > The CXL driver uses both functions phys_to_target_node() and > > memory_add_physaddr_to_nid(). The x86 architecture relies on the > > NUMA_KEEP_MEMINFO kernel option to be set. Enable the option for the > > driver accordingly. > > > > Suggested-by: Dan Williams <dan.j.williams@intel.com> > > Signed-off-by: Robert Richter <rrichter@amd.com> > > --- > > drivers/cxl/Kconfig | 1 + > > 1 file changed, 1 insertion(+) > > > > diff --git a/drivers/cxl/Kconfig b/drivers/cxl/Kconfig > > index 67998dbd1d46..6140b3529a29 100644 > > --- a/drivers/cxl/Kconfig > > +++ b/drivers/cxl/Kconfig > > @@ -6,6 +6,7 @@ menuconfig CXL_BUS > > select FW_UPLOAD > > select PCI_DOE > > select FIRMWARE_TABLE > > + select NUMA_KEEP_MEMINFO if (NUMA && X86) > > help > > CXL is a bus that is electrically compatible with PCI Express, but > > layers three protocols on that signalling (CXL.io, CXL.cache, and > > -- > > 2.39.2 > > > >
diff --git a/drivers/cxl/Kconfig b/drivers/cxl/Kconfig index 67998dbd1d46..6140b3529a29 100644 --- a/drivers/cxl/Kconfig +++ b/drivers/cxl/Kconfig @@ -6,6 +6,7 @@ menuconfig CXL_BUS select FW_UPLOAD select PCI_DOE select FIRMWARE_TABLE + select NUMA_KEEP_MEMINFO if (NUMA && X86) help CXL is a bus that is electrically compatible with PCI Express, but layers three protocols on that signalling (CXL.io, CXL.cache, and