| Message ID | 20230903-cxl-cel-fix-v1-1-e260c9467be3@intel.com |
|---|---|
| State | Accepted |
| Commit | d2f706058826b803f5b9dc3f6d4c213ae0c54eb9 |
| Headers | show |
| Series | cxl/mbox: Fix CEL logic for poison and security commands | expand |
On Sun, 03 Sep 2023 14:42:58 -0700 Ira Weiny <ira.weiny@intel.com> wrote: > The following debug output was observed while testing CXL > > cxl_core:cxl_walk_cel:721: cxl_mock_mem cxl_mem.0: Opcode 0x4300 unsupported by driver > > opcode 0x4300 (Get Poison) is supported by the driver and the mock > device supports it. The logic should be checking that the opcode is > both not poison and not security. > > Fix the logic to allow poison and security commands. > > Fixes: ad64f5952ce3 ("cxl/memdev: Only show sanitize sysfs files when supported") > Cc: Davidlohr Bueso <dave@stgolabs.net> > Signed-off-by: Ira Weiny <ira.weiny@intel.com> Makes sense. Acked-by: Jonathan Cameron <Jonathan.Cameron@huawei.com> > --- > drivers/cxl/core/mbox.c | 4 ++-- > 1 file changed, 2 insertions(+), 2 deletions(-) > > diff --git a/drivers/cxl/core/mbox.c b/drivers/cxl/core/mbox.c > index ca60bb8114f2..b315bdab9197 100644 > --- a/drivers/cxl/core/mbox.c > +++ b/drivers/cxl/core/mbox.c > @@ -716,8 +716,8 @@ static void cxl_walk_cel(struct cxl_memdev_state *mds, size_t size, u8 *cel) > u16 opcode = le16_to_cpu(cel_entry[i].opcode); > struct cxl_mem_command *cmd = cxl_mem_find_command(opcode); > > - if (!cmd && (!cxl_is_poison_command(opcode) || > - !cxl_is_security_command(opcode))) { > + if (!cmd && !cxl_is_poison_command(opcode) && > + !cxl_is_security_command(opcode)) { > dev_dbg(dev, > "Opcode 0x%04x unsupported by driver\n", opcode); > continue; > > --- > base-commit: 1c59d383390f970b891b503b7f79b63a02db2ec5 > change-id: 20230903-cxl-cel-fix-9da269bf0f21 > > Best regards,
On Sun, 03 Sep 2023, Ira Weiny wrote: >The following debug output was observed while testing CXL > >cxl_core:cxl_walk_cel:721: cxl_mock_mem cxl_mem.0: Opcode 0x4300 unsupported by driver > >opcode 0x4300 (Get Poison) is supported by the driver and the mock >device supports it. The logic should be checking that the opcode is >both not poison and not security. > >Fix the logic to allow poison and security commands. Hmm yeah sorry about that. Reviewed-by: Davidlohr Bueso <dave@stgolabs.net> > >Fixes: ad64f5952ce3 ("cxl/memdev: Only show sanitize sysfs files when supported") >Cc: Davidlohr Bueso <dave@stgolabs.net> >Signed-off-by: Ira Weiny <ira.weiny@intel.com> >--- > drivers/cxl/core/mbox.c | 4 ++-- > 1 file changed, 2 insertions(+), 2 deletions(-) > >diff --git a/drivers/cxl/core/mbox.c b/drivers/cxl/core/mbox.c >index ca60bb8114f2..b315bdab9197 100644 >--- a/drivers/cxl/core/mbox.c >+++ b/drivers/cxl/core/mbox.c >@@ -716,8 +716,8 @@ static void cxl_walk_cel(struct cxl_memdev_state *mds, size_t size, u8 *cel) > u16 opcode = le16_to_cpu(cel_entry[i].opcode); > struct cxl_mem_command *cmd = cxl_mem_find_command(opcode); > >- if (!cmd && (!cxl_is_poison_command(opcode) || >- !cxl_is_security_command(opcode))) { >+ if (!cmd && !cxl_is_poison_command(opcode) && >+ !cxl_is_security_command(opcode)) { > dev_dbg(dev, > "Opcode 0x%04x unsupported by driver\n", opcode); > continue; > >--- >base-commit: 1c59d383390f970b891b503b7f79b63a02db2ec5 >change-id: 20230903-cxl-cel-fix-9da269bf0f21 > >Best regards, >-- >Ira Weiny <ira.weiny@intel.com> >
Ira Weiny wrote: > The following debug output was observed while testing CXL > > cxl_core:cxl_walk_cel:721: cxl_mock_mem cxl_mem.0: Opcode 0x4300 unsupported by driver > > opcode 0x4300 (Get Poison) is supported by the driver and the mock > device supports it. The logic should be checking that the opcode is > both not poison and not security. > > Fix the logic to allow poison and security commands. > > Fixes: ad64f5952ce3 ("cxl/memdev: Only show sanitize sysfs files when supported") > Cc: Davidlohr Bueso <dave@stgolabs.net> > Signed-off-by: Ira Weiny <ira.weiny@intel.com> > --- > drivers/cxl/core/mbox.c | 4 ++-- > 1 file changed, 2 insertions(+), 2 deletions(-) > > diff --git a/drivers/cxl/core/mbox.c b/drivers/cxl/core/mbox.c > index ca60bb8114f2..b315bdab9197 100644 > --- a/drivers/cxl/core/mbox.c > +++ b/drivers/cxl/core/mbox.c > @@ -716,8 +716,8 @@ static void cxl_walk_cel(struct cxl_memdev_state *mds, size_t size, u8 *cel) > u16 opcode = le16_to_cpu(cel_entry[i].opcode); > struct cxl_mem_command *cmd = cxl_mem_find_command(opcode); > > - if (!cmd && (!cxl_is_poison_command(opcode) || > - !cxl_is_security_command(opcode))) { > + if (!cmd && !cxl_is_poison_command(opcode) && > + !cxl_is_security_command(opcode)) { Given that this is going to be a recurring pattern to add optional command support, I am not a fan of continuing to expand this boolean algebra eye exam. How about the following to only do the validation check and enabling in one place: diff --git a/drivers/cxl/core/mbox.c b/drivers/cxl/core/mbox.c index ca60bb8114f2..4df4f614f490 100644 --- a/drivers/cxl/core/mbox.c +++ b/drivers/cxl/core/mbox.c @@ -715,24 +715,25 @@ static void cxl_walk_cel(struct cxl_memdev_state *mds, size_t size, u8 *cel) for (i = 0; i < cel_entries; i++) { u16 opcode = le16_to_cpu(cel_entry[i].opcode); struct cxl_mem_command *cmd = cxl_mem_find_command(opcode); + int enabled = 0; - if (!cmd && (!cxl_is_poison_command(opcode) || - !cxl_is_security_command(opcode))) { - dev_dbg(dev, - "Opcode 0x%04x unsupported by driver\n", opcode); - continue; - } - - if (cmd) + if (cmd) { set_bit(cmd->info.id, mds->enabled_cmds); + enabled++; + } - if (cxl_is_poison_command(opcode)) + if (cxl_is_poison_command(opcode)) { cxl_set_poison_cmd_enabled(&mds->poison, opcode); + enabled++; + } - if (cxl_is_security_command(opcode)) + if (cxl_is_security_command(opcode)) { cxl_set_security_cmd_enabled(&mds->security, opcode); + enabled++; + } - dev_dbg(dev, "Opcode 0x%04x enabled\n", opcode); + dev_dbg(dev, "Opcode 0x%04x %s\n", opcode, + enabled ? "enabled" : "unsupported by driver"); } }
Dan Williams wrote: > Ira Weiny wrote: > > The following debug output was observed while testing CXL > > > > cxl_core:cxl_walk_cel:721: cxl_mock_mem cxl_mem.0: Opcode 0x4300 unsupported by driver > > > > opcode 0x4300 (Get Poison) is supported by the driver and the mock > > device supports it. The logic should be checking that the opcode is > > both not poison and not security. > > > > Fix the logic to allow poison and security commands. > > > > Fixes: ad64f5952ce3 ("cxl/memdev: Only show sanitize sysfs files when supported") > > Cc: Davidlohr Bueso <dave@stgolabs.net> > > Signed-off-by: Ira Weiny <ira.weiny@intel.com> > > --- > > drivers/cxl/core/mbox.c | 4 ++-- > > 1 file changed, 2 insertions(+), 2 deletions(-) > > > > diff --git a/drivers/cxl/core/mbox.c b/drivers/cxl/core/mbox.c > > index ca60bb8114f2..b315bdab9197 100644 > > --- a/drivers/cxl/core/mbox.c > > +++ b/drivers/cxl/core/mbox.c > > @@ -716,8 +716,8 @@ static void cxl_walk_cel(struct cxl_memdev_state *mds, size_t size, u8 *cel) > > u16 opcode = le16_to_cpu(cel_entry[i].opcode); > > struct cxl_mem_command *cmd = cxl_mem_find_command(opcode); > > > > - if (!cmd && (!cxl_is_poison_command(opcode) || > > - !cxl_is_security_command(opcode))) { > > + if (!cmd && !cxl_is_poison_command(opcode) && > > + !cxl_is_security_command(opcode)) { > > Given that this is going to be a recurring pattern to add optional > command support, I am not a fan of continuing to expand this boolean > algebra eye exam. > > How about the following to only do the validation check and enabling in > one place: I like it. Reviewed-by: Ira Weiny <ira.weiny@intel.com> > > > diff --git a/drivers/cxl/core/mbox.c b/drivers/cxl/core/mbox.c > index ca60bb8114f2..4df4f614f490 100644 > --- a/drivers/cxl/core/mbox.c > +++ b/drivers/cxl/core/mbox.c > @@ -715,24 +715,25 @@ static void cxl_walk_cel(struct cxl_memdev_state *mds, size_t size, u8 *cel) > for (i = 0; i < cel_entries; i++) { > u16 opcode = le16_to_cpu(cel_entry[i].opcode); > struct cxl_mem_command *cmd = cxl_mem_find_command(opcode); > + int enabled = 0; > > - if (!cmd && (!cxl_is_poison_command(opcode) || > - !cxl_is_security_command(opcode))) { > - dev_dbg(dev, > - "Opcode 0x%04x unsupported by driver\n", opcode); > - continue; > - } > - > - if (cmd) > + if (cmd) { > set_bit(cmd->info.id, mds->enabled_cmds); > + enabled++; > + } > > - if (cxl_is_poison_command(opcode)) > + if (cxl_is_poison_command(opcode)) { > cxl_set_poison_cmd_enabled(&mds->poison, opcode); > + enabled++; > + } > > - if (cxl_is_security_command(opcode)) > + if (cxl_is_security_command(opcode)) { > cxl_set_security_cmd_enabled(&mds->security, opcode); > + enabled++; > + } > > - dev_dbg(dev, "Opcode 0x%04x enabled\n", opcode); > + dev_dbg(dev, "Opcode 0x%04x %s\n", opcode, > + enabled ? "enabled" : "unsupported by driver"); > } > } >
On Tue, 12 Sep 2023, Dan Williams wrote: >Ira Weiny wrote: >> The following debug output was observed while testing CXL >> >> cxl_core:cxl_walk_cel:721: cxl_mock_mem cxl_mem.0: Opcode 0x4300 unsupported by driver >> >> opcode 0x4300 (Get Poison) is supported by the driver and the mock >> device supports it. The logic should be checking that the opcode is >> both not poison and not security. >> >> Fix the logic to allow poison and security commands. >> >> Fixes: ad64f5952ce3 ("cxl/memdev: Only show sanitize sysfs files when supported") >> Cc: Davidlohr Bueso <dave@stgolabs.net> >> Signed-off-by: Ira Weiny <ira.weiny@intel.com> >> --- >> drivers/cxl/core/mbox.c | 4 ++-- >> 1 file changed, 2 insertions(+), 2 deletions(-) >> >> diff --git a/drivers/cxl/core/mbox.c b/drivers/cxl/core/mbox.c >> index ca60bb8114f2..b315bdab9197 100644 >> --- a/drivers/cxl/core/mbox.c >> +++ b/drivers/cxl/core/mbox.c >> @@ -716,8 +716,8 @@ static void cxl_walk_cel(struct cxl_memdev_state *mds, size_t size, u8 *cel) >> u16 opcode = le16_to_cpu(cel_entry[i].opcode); >> struct cxl_mem_command *cmd = cxl_mem_find_command(opcode); >> >> - if (!cmd && (!cxl_is_poison_command(opcode) || >> - !cxl_is_security_command(opcode))) { >> + if (!cmd && !cxl_is_poison_command(opcode) && >> + !cxl_is_security_command(opcode)) { > >Given that this is going to be a recurring pattern to add optional >command support, I am not a fan of continuing to expand this boolean >algebra eye exam. > >How about the following to only do the validation check and enabling in >one place: Agreed. Reviewed-by: Davidlohr Bueso <dave@stgolabs.net> ... this also is a reminder of the need for regression testing/CI. Thanks, Davidlohr
diff --git a/drivers/cxl/core/mbox.c b/drivers/cxl/core/mbox.c index ca60bb8114f2..b315bdab9197 100644 --- a/drivers/cxl/core/mbox.c +++ b/drivers/cxl/core/mbox.c @@ -716,8 +716,8 @@ static void cxl_walk_cel(struct cxl_memdev_state *mds, size_t size, u8 *cel) u16 opcode = le16_to_cpu(cel_entry[i].opcode); struct cxl_mem_command *cmd = cxl_mem_find_command(opcode); - if (!cmd && (!cxl_is_poison_command(opcode) || - !cxl_is_security_command(opcode))) { + if (!cmd && !cxl_is_poison_command(opcode) && + !cxl_is_security_command(opcode)) { dev_dbg(dev, "Opcode 0x%04x unsupported by driver\n", opcode); continue;
The following debug output was observed while testing CXL cxl_core:cxl_walk_cel:721: cxl_mock_mem cxl_mem.0: Opcode 0x4300 unsupported by driver opcode 0x4300 (Get Poison) is supported by the driver and the mock device supports it. The logic should be checking that the opcode is both not poison and not security. Fix the logic to allow poison and security commands. Fixes: ad64f5952ce3 ("cxl/memdev: Only show sanitize sysfs files when supported") Cc: Davidlohr Bueso <dave@stgolabs.net> Signed-off-by: Ira Weiny <ira.weiny@intel.com> --- drivers/cxl/core/mbox.c | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) --- base-commit: 1c59d383390f970b891b503b7f79b63a02db2ec5 change-id: 20230903-cxl-cel-fix-9da269bf0f21 Best regards,