Skip to contentSkip to conversation

create-spdx-3.0: make create_spdx depend on CVE_STATUS

PATCH In review Peter Marko 3 participants 6 messages Mar 31, 2025

lore
Share & tools
lore
#0 lore raw

PATCH 1/2 create-spdx-3.0: make create_spdx depend on CVE_STATUS

From: Peter Marko <[email protected]>

SPDX3 classes now add CVE_STATUS information to spdx. So the task needs to be rebuilt when CVE_STATUS changes.

This is true since commit 1ff496546279d8a97df5ec475007cfb095c2a0bc

---
 meta/classes/create-spdx-3.0.bbclass | 1 +
 1 file changed, 1 insertion(+)
meta/classes/create-spdx-3.0.bbclass +1 -0
index 044517d9f7..235bd804bd 100644
@@ -138,6 +138,7 @@ python do_create_spdx() {
     oe.spdx30_tasks.create_spdx(d)
 }
 do_create_spdx[vardeps] += "\
+    CVE_STATUS \
     SPDX_INCLUDE_BITBAKE_PARENT_BUILD \
     SPDX_PACKAGE_ADDITIONAL_PURPOSE \
     SPDX_PROFILES \
#1 lore raw

PATCH 2/2 spdx30: handle Unknown CVE_STATUS

From: Peter Marko <[email protected]>

CVE_STATUS can be also "Unknown" since oe-core commit d25f1817752bc8a84c40dcbef75f7559801ce15e

When this status type is used, build fails with e.g.
ERROR: openssl-3.4.1-r0 do_create_spdx: Unknown CVE-2025-0001 status 'Unknown'

Since this is now a valid status, it needs to be handled.
It cannot be mapped to any VEX status (see below), so just skip it. Possible VEX statuses are: NOT AFFECTED, AFFECTED, FIXED, and UNDER INVESTIGATION.

cc Marta Rybczynska <[email protected]>
---
 meta/lib/oe/spdx30_tasks.py | 2 ++
 1 file changed, 2 insertions(+)
meta/lib/oe/spdx30_tasks.py +2 -0
index 52329760b6..ba965821f8 100644
@@ -724,6 +724,8 @@ def create_spdx(d):
                                 )
                         else:
                             bb.fatal(f"Unknown detail '{detail}' for ignored {cve}")
+                    elif status == "Unknown":
+                        bb.note(f"Skipping {cve} with status 'Unknown'")
                     else:
                         bb.fatal(f"Unknown {cve} status '{status}'")
 
Reviewed-by on v1 1/2
#2 lore raw
In reply to Peter Marko
Quoted 19 lines · Peter Marko

From: Peter Marko <[email protected]>

SPDX3 classes now add CVE_STATUS information to spdx. So the task needs to be rebuilt when CVE_STATUS changes.

This is true since commit 1ff496546279d8a97df5ec475007cfb095c2a0bc

---
 meta/classes/create-spdx-3.0.bbclass | 1 +
 1 file changed, 1 insertion(+)
meta/classes/create-spdx-3.0.bbclass +0 -0
index 044517d9f7..235bd804bd 100644

meta/classes/create-spdx-3.0.bbclass python do_create_spdx() { · line 141

Quoted 4 lines
meta/classes/create-spdx-3.0.bbclass +1 -0
@@ -138,6 +138,7 @@ python do_create_spdx() {
     oe.spdx30_tasks.create_spdx(d)
 }
 do_create_spdx[vardeps] += "\
+    CVE_STATUS \

Seems fine to me, unless RP thinks we should add cve_check to BBIMPORTS instead?

Quoted 3 lines
meta/classes/create-spdx-3.0.bbclass +0 -0
     SPDX_INCLUDE_BITBAKE_PARENT_BUILD \
     SPDX_PACKAGE_ADDITIONAL_PURPOSE \
     SPDX_PROFILES \
#3 lore raw
Quoted 16 lines · Joshua Watt
Quoted 14 lines · Peter Marko

From: Peter Marko <[email protected]>

SPDX3 classes now add CVE_STATUS information to spdx. So the task needs to be rebuilt when CVE_STATUS changes.

This is true since commit 1ff496546279d8a97df5ec475007cfb095c2a0bc

---
 meta/classes/create-spdx-3.0.bbclass | 1 +

 1 file changed, 1 insertion(+)
Quoted 13 lines

meta/classes/create-spdx-3.0.bbclass python do_create_spdx() { · line 141

Quoted 10 lines
meta/classes/create-spdx-3.0.bbclass +1 -0
index 044517d9f7..235bd804bd 100644
@@ -138,6 +138,7 @@ python do_create_spdx() {
      oe.spdx30_tasks.create_spdx(d)
  }
  do_create_spdx[vardeps] += "\
+    CVE_STATUS \

Seems fine to me, unless RP thinks we should add cve_check to BBIMPORTS instead?

That would potentially see off other problems in the future. Is there a reason we wouldn't want to do that?

Cheers,

Richard

#4 lore raw
Quoted 32 lines · Richard Purdie
Quoted 16 lines · Joshua Watt
Quoted 14 lines · Peter Marko

From: Peter Marko <[email protected]>

SPDX3 classes now add CVE_STATUS information to spdx. So the task needs to be rebuilt when CVE_STATUS changes.

This is true since commit 1ff496546279d8a97df5ec475007cfb095c2a0bc

---
 meta/classes/create-spdx-3.0.bbclass | 1 +
 1 file changed, 1 insertion(+)
Quoted 13 lines

meta/classes/create-spdx-3.0.bbclass python do_create_spdx() { · line 141

Quoted 10 lines
meta/classes/create-spdx-3.0.bbclass +1 -0
index 044517d9f7..235bd804bd 100644
@@ -138,6 +138,7 @@ python do_create_spdx() {
     oe.spdx30_tasks.create_spdx(d)
 }
 do_create_spdx[vardeps] += "\
+    CVE_STATUS \

Seems fine to me, unless RP thinks we should add cve_check to BBIMPORTS instead?

Quoted 2 lines

That would potentially see off other problems in the future. Is there a reason we wouldn't want to do that?

I can't think of one, as long as it actually works :). I'm testing it now and I'll send the patch if works

Quoted 4 lines

Cheers,

Richard

#5 lore raw
Quoted 35 lines · Joshua Watt
Quoted 32 lines · Richard Purdie
Quoted 29 lines · Joshua Watt
Quoted 20 lines · Peter Marko

From: Peter Marko <[email protected]>

SPDX3 classes now add CVE_STATUS information to spdx. So the task needs to be rebuilt when CVE_STATUS changes.

This is true since commit 1ff496546279d8a97df5ec475007cfb095c2a0bc

---
 meta/classes/create-spdx-3.0.bbclass | 1 +
 1 file changed, 1 insertion(+)
meta/classes/create-spdx-3.0.bbclass +0 -0
index 044517d9f7..235bd804bd 100644

meta/classes/create-spdx-3.0.bbclass python do_create_spdx() { · line 141

Quoted 4 lines
meta/classes/create-spdx-3.0.bbclass +1 -0
@@ -138,6 +138,7 @@ python do_create_spdx() {
     oe.spdx30_tasks.create_spdx(d)
 }
 do_create_spdx[vardeps] += "\
+    CVE_STATUS \

Seems fine to me, unless RP thinks we should add cve_check to BBIMPORTS instead?

Quoted 5 lines
Quoted 2 lines

That would potentially see off other problems in the future. Is there a reason we wouldn't want to do that?

I can't think of one, as long as it actually works :). I'm testing it now and I'll send the patch if works

Just to follow up on this; I think we can fix this more systemically in bitbake and/or make improvements to make it easier to specify these types of dependencies in the actual code that references them (see jpew/var-flag-deps on poky-contrib for some prototypes), so I think it would be better to take that approach than this one (at least on master, this is likely better for backport fixes)

Quoted 5 lines
Quoted 5 lines

Cheers,

Richard