* [OE-core][kirkstone 0/8] Patch review
@ 2024-02-12 13:54 Steve Sakoman
2024-02-12 13:54 ` [OE-core][kirkstone 1/8] libxml2: Fix for CVE-2024-25062 Steve Sakoman
` (7 more replies)
0 siblings, 8 replies; 33+ messages in thread
From: Steve Sakoman @ 2024-02-12 13:54 UTC (permalink / raw)
To: openembedded-core
Please review this set of changes for kirkstone and have comments back by
end of day Wednesday, February 14
Passed a-full on autobuilder:
https://autobuilder.yoctoproject.org/typhoon/#/builders/83/builds/6556
The following changes since commit 2bdae590ab20dc4518ba247c903060fa67ed0fc4:
openssl: Upgrade 3.0.12 -> 3.0.13 (2024-02-05 03:56:38 -1000)
are available in the Git repository at:
https://git.openembedded.org/openembedded-core-contrib stable/kirkstone-nut
https://git.openembedded.org/openembedded-core-contrib/log/?h=stable/kirkstone-nut
Archana Polampalli (1):
curl: Fix CVE-2023-46219
Bruce Ashfield (1):
kernel: fix localversion in v6.3+
Jermain Horsman (1):
systemd: Only add myhostname to nsswitch.conf if in PACKAGECONFIG
Kai Kang (1):
ghostscript: correct LICENSE with AGPLv3
Narpat Mali (1):
python3-pycryptodome: Fix CVE-2023-52323
Soumya Sambu (2):
go: Fix CVE-2023-45285 and CVE-2023-45287
libgit2: Fix CVE-2024-24575 and CVE-2024-24577
Vijay Anusuri (1):
libxml2: Fix for CVE-2024-25062
meta/classes/kernel-arch.bbclass | 7 +
meta/classes/kernel.bbclass | 10 +-
.../libxml/libxml2/CVE-2024-25062.patch | 33 +
meta/recipes-core/libxml/libxml2_2.9.14.bb | 1 +
meta/recipes-core/systemd/systemd_250.5.bb | 16 +-
meta/recipes-devtools/go/go-1.17.13.inc | 2 +
.../go/go-1.20/CVE-2023-45285.patch | 110 ++
.../go/go-1.20/CVE-2023-45287.patch | 1695 +++++++++++++++++
.../python3-pycryptodome/CVE-2023-52323.patch | 436 +++++
.../python/python3-pycryptodome_3.14.1.bb | 1 +
.../CVE-2023-52323.patch | 436 +++++
.../python/python3-pycryptodomex_3.14.1.bb | 2 +
.../ghostscript/ghostscript_9.55.0.bb | 2 +-
.../curl/curl/CVE-2023-46219-0001.patch | 42 +
.../curl/curl/CVE-2023-46219-0002.patch | 133 ++
.../curl/curl/CVE-2023-46219-0003.patch | 81 +
meta/recipes-support/curl/curl_7.82.0.bb | 3 +
.../libgit2/libgit2/CVE-2024-24575.patch | 56 +
.../libgit2/libgit2/CVE-2024-24577.patch | 52 +
meta/recipes-support/libgit2/libgit2_1.4.5.bb | 5 +-
20 files changed, 3113 insertions(+), 10 deletions(-)
create mode 100644 meta/recipes-core/libxml/libxml2/CVE-2024-25062.patch
create mode 100644 meta/recipes-devtools/go/go-1.20/CVE-2023-45285.patch
create mode 100644 meta/recipes-devtools/go/go-1.20/CVE-2023-45287.patch
create mode 100644 meta/recipes-devtools/python/python3-pycryptodome/CVE-2023-52323.patch
create mode 100644 meta/recipes-devtools/python/python3-pycryptodomex/CVE-2023-52323.patch
create mode 100644 meta/recipes-support/curl/curl/CVE-2023-46219-0001.patch
create mode 100644 meta/recipes-support/curl/curl/CVE-2023-46219-0002.patch
create mode 100644 meta/recipes-support/curl/curl/CVE-2023-46219-0003.patch
create mode 100644 meta/recipes-support/libgit2/libgit2/CVE-2024-24575.patch
create mode 100644 meta/recipes-support/libgit2/libgit2/CVE-2024-24577.patch
--
2.34.1
^ permalink raw reply [flat|nested] 33+ messages in thread* [OE-core][kirkstone 1/8] libxml2: Fix for CVE-2024-25062 2024-02-12 13:54 [OE-core][kirkstone 0/8] Patch review Steve Sakoman @ 2024-02-12 13:54 ` Steve Sakoman 2024-02-12 13:54 ` [OE-core][kirkstone 2/8] go: Fix CVE-2023-45285 and CVE-2023-45287 Steve Sakoman ` (6 subsequent siblings) 7 siblings, 0 replies; 33+ messages in thread From: Steve Sakoman @ 2024-02-12 13:54 UTC (permalink / raw) To: openembedded-core From: Vijay Anusuri <vanusuri@mvista.com> Upstream-Status: Backport [https://gitlab.gnome.org/GNOME/libxml2/-/commit/2b0aac140d739905c7848a42efc60bfe783a39b7] Signed-off-by: Vijay Anusuri <vanusuri@mvista.com> Signed-off-by: Steve Sakoman <steve@sakoman.com> --- .../libxml/libxml2/CVE-2024-25062.patch | 33 +++++++++++++++++++ meta/recipes-core/libxml/libxml2_2.9.14.bb | 1 + 2 files changed, 34 insertions(+) create mode 100644 meta/recipes-core/libxml/libxml2/CVE-2024-25062.patch diff --git a/meta/recipes-core/libxml/libxml2/CVE-2024-25062.patch b/meta/recipes-core/libxml/libxml2/CVE-2024-25062.patch new file mode 100644 index 0000000000..5365d5546a --- /dev/null +++ b/meta/recipes-core/libxml/libxml2/CVE-2024-25062.patch @@ -0,0 +1,33 @@ +From 2b0aac140d739905c7848a42efc60bfe783a39b7 Mon Sep 17 00:00:00 2001 +From: Nick Wellnhofer <wellnhofer@aevum.de> +Date: Sat, 14 Oct 2023 22:45:54 +0200 +Subject: [PATCH] [CVE-2024-25062] xmlreader: Don't expand XIncludes when + backtracking + +Fixes a use-after-free if XML Reader if used with DTD validation and +XInclude expansion. + +Fixes #604. + +Upstream-Status: Backport [https://gitlab.gnome.org/GNOME/libxml2/-/commit/2b0aac140d739905c7848a42efc60bfe783a39b7] +CVE: CVE-2024-25062 +Signed-off-by: Vijay Anusuri <vanusuri@mvista.com> +--- + xmlreader.c | 1 + + 1 file changed, 1 insertion(+) + +diff --git a/xmlreader.c b/xmlreader.c +index 979385a13..fefd68e0b 100644 +--- a/xmlreader.c ++++ b/xmlreader.c +@@ -1443,6 +1443,7 @@ node_found: + * Handle XInclude if asked for + */ + if ((reader->xinclude) && (reader->in_xinclude == 0) && ++ (reader->state != XML_TEXTREADER_BACKTRACK) && + (reader->node != NULL) && + (reader->node->type == XML_ELEMENT_NODE) && + (reader->node->ns != NULL) && +-- +GitLab + diff --git a/meta/recipes-core/libxml/libxml2_2.9.14.bb b/meta/recipes-core/libxml/libxml2_2.9.14.bb index 533a6dae01..2b7e9999d9 100644 --- a/meta/recipes-core/libxml/libxml2_2.9.14.bb +++ b/meta/recipes-core/libxml/libxml2_2.9.14.bb @@ -31,6 +31,7 @@ SRC_URI += "http://www.w3.org/XML/Test/xmlts20080827.tar;subdir=${BP};name=testt file://CVE-2023-39615-0002.patch \ file://CVE-2023-45322-1.patch \ file://CVE-2023-45322-2.patch \ + file://CVE-2024-25062.patch \ " SRC_URI[archive.sha256sum] = "60d74a257d1ccec0475e749cba2f21559e48139efba6ff28224357c7c798dfee" -- 2.34.1 ^ permalink raw reply related [flat|nested] 33+ messages in thread
* [OE-core][kirkstone 2/8] go: Fix CVE-2023-45285 and CVE-2023-45287 2024-02-12 13:54 [OE-core][kirkstone 0/8] Patch review Steve Sakoman 2024-02-12 13:54 ` [OE-core][kirkstone 1/8] libxml2: Fix for CVE-2024-25062 Steve Sakoman @ 2024-02-12 13:54 ` Steve Sakoman 2024-02-12 13:54 ` [OE-core][kirkstone 3/8] curl: Fix CVE-2023-46219 Steve Sakoman ` (5 subsequent siblings) 7 siblings, 0 replies; 33+ messages in thread From: Steve Sakoman @ 2024-02-12 13:54 UTC (permalink / raw) To: openembedded-core From: Soumya Sambu <soumya.sambu@windriver.com> CVE-2023-45285: Using go get to fetch a module with the ".git" suffix may unexpectedly fallback to the insecure "git://" protocol if the module is unavailable via the secure "https://" and "git+ssh://" protocols, even if GOINSECURE is not set for said module. This only affects users who are not using the module proxy and are fetching modules directly (i.e. GOPROXY=off). CVE-2023-45287: Before Go 1.20, the RSA based TLS key exchanges used the math/big library, which is not constant time. RSA blinding was applied to prevent timing attacks, but analysis shows this may not have been fully effective. In particular it appears as if the removal of PKCS#1 padding may leak timing information, which in turn could be used to recover session key bits. In Go 1.20, the crypto/tls library switched to a fully constant time RSA implementation, which we do not believe exhibits any timing side channels. References: https://nvd.nist.gov/vuln/detail/CVE-2023-45285 https://nvd.nist.gov/vuln/detail/CVE-2023-45287 https://security-tracker.debian.org/tracker/CVE-2023-45285 https://security-tracker.debian.org/tracker/CVE-2023-45287 Signed-off-by: Soumya Sambu <soumya.sambu@windriver.com> Signed-off-by: Steve Sakoman <steve@sakoman.com> --- meta/recipes-devtools/go/go-1.17.13.inc | 2 + .../go/go-1.20/CVE-2023-45285.patch | 110 ++ .../go/go-1.20/CVE-2023-45287.patch | 1695 +++++++++++++++++ 3 files changed, 1807 insertions(+) create mode 100644 meta/recipes-devtools/go/go-1.20/CVE-2023-45285.patch create mode 100644 meta/recipes-devtools/go/go-1.20/CVE-2023-45287.patch diff --git a/meta/recipes-devtools/go/go-1.17.13.inc b/meta/recipes-devtools/go/go-1.17.13.inc index 95c4461d3e..6fc07bb910 100644 --- a/meta/recipes-devtools/go/go-1.17.13.inc +++ b/meta/recipes-devtools/go/go-1.17.13.inc @@ -48,6 +48,8 @@ SRC_URI += "\ file://CVE-2023-39319.patch \ file://CVE-2023-39318.patch \ file://CVE-2023-39326.patch \ + file://CVE-2023-45285.patch \ + file://CVE-2023-45287.patch \ " SRC_URI[main.sha256sum] = "a1a48b23afb206f95e7bbaa9b898d965f90826f6f1d1fc0c1d784ada0cd300fd" diff --git a/meta/recipes-devtools/go/go-1.20/CVE-2023-45285.patch b/meta/recipes-devtools/go/go-1.20/CVE-2023-45285.patch new file mode 100644 index 0000000000..0459ae0a1a --- /dev/null +++ b/meta/recipes-devtools/go/go-1.20/CVE-2023-45285.patch @@ -0,0 +1,110 @@ +From 46bc33819ac86a9596b8059235842f0e0c7469bd Mon Sep 17 00:00:00 2001 +From: Bryan C. Mills <bcmills@google.com> +Date: Thu, 2 Nov 2023 15:06:35 -0400 +Subject: [PATCH] cmd/go/internal/vcs: error out if the requested repo does not + support a secure protocol + +Updates #63845. +Fixes #63972. + +Change-Id: If86d6b13d3b55877b35c087112bd76388c9404b8 +Reviewed-on: https://go-review.googlesource.com/c/go/+/539321 +Reviewed-by: Michael Matloob <matloob@golang.org> +LUCI-TryBot-Result: Go LUCI <golang-scoped@luci-project-accounts.iam.gserviceaccount.com> +Reviewed-by: Roland Shoemaker <roland@golang.org> +Auto-Submit: Bryan Mills <bcmills@google.com> +(cherry picked from commit be26ae18caf7ddffca4073333f80d0d9e76483c3) +Reviewed-on: https://go-review.googlesource.com/c/go/+/540335 +Auto-Submit: Dmitri Shuralyov <dmitshur@google.com> +Reviewed-by: Dmitri Shuralyov <dmitshur@google.com> + +CVE: CVE-2023-45285 + +Upstream-Status: Backport [https://github.com/golang/go/commit/46bc33819ac86a9596b8059235842f0e0c7469bd] + +Signed-off-by: Soumya Sambu <soumya.sambu@windriver.com> +--- + src/cmd/go/internal/vcs/vcs.go | 25 +++++++++++++---- + .../script/mod_insecure_issue63845.txt | 28 +++++++++++++++++++ + 2 files changed, 47 insertions(+), 6 deletions(-) + create mode 100644 src/cmd/go/testdata/script/mod_insecure_issue63845.txt + +diff --git a/src/cmd/go/internal/vcs/vcs.go b/src/cmd/go/internal/vcs/vcs.go +index ab42424..0e2882d 100644 +--- a/src/cmd/go/internal/vcs/vcs.go ++++ b/src/cmd/go/internal/vcs/vcs.go +@@ -891,19 +891,32 @@ func repoRootFromVCSPaths(importPath string, security web.SecurityMode, vcsPaths + if !srv.schemelessRepo { + repoURL = match["repo"] + } else { +- scheme := vcs.Scheme[0] // default to first scheme + repo := match["repo"] +- if vcs.PingCmd != "" { +- // If we know how to test schemes, scan to find one. ++ scheme, err := func() (string, error) { + for _, s := range vcs.Scheme { + if security == web.SecureOnly && !vcs.isSecureScheme(s) { + continue + } +- if vcs.Ping(s, repo) == nil { +- scheme = s +- break ++ ++ // If we know how to ping URL schemes for this VCS, ++ // check that this repo works. ++ // Otherwise, default to the first scheme ++ // that meets the requested security level. ++ if vcs.PingCmd == "" { ++ return s, nil ++ } ++ if err := vcs.Ping(s, repo); err == nil { ++ return s, nil + } + } ++ securityFrag := "" ++ if security == web.SecureOnly { ++ securityFrag = "secure " ++ } ++ return "", fmt.Errorf("no %sprotocol found for repository", securityFrag) ++ }() ++ if err != nil { ++ return nil, err + } + repoURL = scheme + "://" + repo + } +diff --git a/src/cmd/go/testdata/script/mod_insecure_issue63845.txt b/src/cmd/go/testdata/script/mod_insecure_issue63845.txt +new file mode 100644 +index 0000000..5fa6a4f +--- /dev/null ++++ b/src/cmd/go/testdata/script/mod_insecure_issue63845.txt +@@ -0,0 +1,28 @@ ++# Regression test for https://go.dev/issue/63845: ++# If 'git ls-remote' fails for all secure protocols, ++# we should fail instead of falling back to an arbitrary protocol. ++# ++# Note that this test does not use the local vcweb test server ++# (vcs-test.golang.org), because the hook for redirecting to that ++# server bypasses the "ping to determine protocol" logic ++# in cmd/go/internal/vcs. ++ ++[!net] skip ++[!git] skip ++[short] skip 'tries to access a nonexistent external Git repo' ++ ++env GOPRIVATE=golang.org ++env CURLOPT_TIMEOUT_MS=100 ++env GIT_SSH_COMMAND=false ++ ++! go get -x golang.org/nonexist.git@latest ++stderr '^git ls-remote https://golang.org/nonexist$' ++stderr '^git ls-remote git\+ssh://golang.org/nonexist' ++stderr '^git ls-remote ssh://golang.org/nonexist$' ++! stderr 'git://' ++stderr '^go: golang.org/nonexist.git@latest: no secure protocol found for repository$' ++ ++-- go.mod -- ++module example ++ ++go 1.19 +-- +2.40.0 diff --git a/meta/recipes-devtools/go/go-1.20/CVE-2023-45287.patch b/meta/recipes-devtools/go/go-1.20/CVE-2023-45287.patch new file mode 100644 index 0000000000..477e3c98ee --- /dev/null +++ b/meta/recipes-devtools/go/go-1.20/CVE-2023-45287.patch @@ -0,0 +1,1695 @@ +From 8a81fdf165facdcefa06531de5af98a4db343035 Mon Sep 17 00:00:00 2001 +From: Lúcás Meier <cronokirby@gmail.com> +Date: Tue Jun 8 21:36:06 2021 +0200 +Subject: [PATCH] crypto/rsa: replace big.Int for encryption and decryption + +Infamously, big.Int does not provide constant-time arithmetic, making +its use in cryptographic code quite tricky. RSA uses big.Int +pervasively, in its public API, for key generation, precomputation, and +for encryption and decryption. This is a known problem. One mitigation, +blinding, is already in place during decryption. This helps mitigate the +very leaky exponentiation operation. Because big.Int is fundamentally +not constant-time, it's unfortunately difficult to guarantee that +mitigations like these are completely effective. + +This patch removes the use of big.Int for encryption and decryption, +replacing it with an internal nat type instead. Signing and verification +are also affected, because they depend on encryption and decryption. + +Overall, this patch degrades performance by 55% for private key +operations, and 4-5x for (much faster) public key operations. +(Signatures do both, so the slowdown is worse than decryption.) + +name old time/op new time/op delta +DecryptPKCS1v15/2048-8 1.50ms ± 0% 2.34ms ± 0% +56.44% (p=0.000 n=8+10) +DecryptPKCS1v15/3072-8 4.40ms ± 0% 6.79ms ± 0% +54.33% (p=0.000 n=10+9) +DecryptPKCS1v15/4096-8 9.31ms ± 0% 15.14ms ± 0% +62.60% (p=0.000 n=10+10) +EncryptPKCS1v15/2048-8 8.16µs ± 0% 355.58µs ± 0% +4258.90% (p=0.000 n=10+9) +DecryptOAEP/2048-8 1.50ms ± 0% 2.34ms ± 0% +55.68% (p=0.000 n=10+9) +EncryptOAEP/2048-8 8.51µs ± 0% 355.95µs ± 0% +4082.75% (p=0.000 n=10+9) +SignPKCS1v15/2048-8 1.51ms ± 0% 2.69ms ± 0% +77.94% (p=0.000 n=10+10) +VerifyPKCS1v15/2048-8 7.25µs ± 0% 354.34µs ± 0% +4789.52% (p=0.000 n=9+9) +SignPSS/2048-8 1.51ms ± 0% 2.70ms ± 0% +78.80% (p=0.000 n=9+10) +VerifyPSS/2048-8 8.27µs ± 1% 355.65µs ± 0% +4199.39% (p=0.000 n=10+10) + +Keep in mind that this is without any assembly at all, and that further +improvements are likely possible. I think having a review of the logic +and the cryptography would be a good idea at this stage, before we +complicate the code too much through optimization. + +The bulk of the work is in nat.go. This introduces two new types: nat, +representing natural numbers, and modulus, representing moduli used in +modular arithmetic. + +A nat has an "announced size", which may be larger than its "true size", +the number of bits needed to represent this number. Operations on a nat +will only ever leak its announced size, never its true size, or other +information about its value. The size of a nat is always clear based on +how its value is set. For example, x.mod(y, m) will make the announced +size of x match that of m, since x is reduced modulo m. + +Operations assume that the announced size of the operands match what's +expected (with a few exceptions). For example, x.modAdd(y, m) assumes +that x and y have the same announced size as m, and that they're reduced +modulo m. + +Nats are represented over unsatured bits.UintSize - 1 bit limbs. This +means that we can't reuse the assembly routines for big.Int, which use +saturated bits.UintSize limbs. The advantage of unsaturated limbs is +that it makes Montgomery multiplication faster, by needing fewer +registers in a hot loop. This makes exponentiation faster, which +consists of many Montgomery multiplications. + +Moduli use nat internally. Unlike nat, the true size of a modulus always +matches its announced size. When creating a modulus, any zero padding is +removed. Moduli will also precompute constants when created, which is +another reason why having a separate type is desirable. + +Updates #20654 + +Co-authored-by: Filippo Valsorda <filippo@golang.org> +Change-Id: I73b61f87d58ab912e80a9644e255d552cbadcced +Reviewed-on: https://go-review.googlesource.com/c/go/+/326012 +Run-TryBot: Filippo Valsorda <filippo@golang.org> +TryBot-Result: Gopher Robot <gobot@golang.org> +Reviewed-by: Roland Shoemaker <roland@golang.org> +Reviewed-by: Joedian Reid <joedian@golang.org> + +CVE: CVE-2023-45287 + +Upstream-Status: Backport [https://github.com/golang/go/commit/8a81fdf165facdcefa06531de5af98a4db343035] + +Signed-off-by: Soumya Sambu <soumya.sambu@windriver.com> +--- + src/crypto/rsa/example_test.go | 21 +- + src/crypto/rsa/nat.go | 626 +++++++++++++++++++++++++++++++++ + src/crypto/rsa/nat_test.go | 384 ++++++++++++++++++++ + src/crypto/rsa/pkcs1v15.go | 47 +-- + src/crypto/rsa/pss.go | 49 ++- + src/crypto/rsa/pss_test.go | 10 +- + src/crypto/rsa/rsa.go | 172 ++++----- + 7 files changed, 1140 insertions(+), 169 deletions(-) + create mode 100644 src/crypto/rsa/nat.go + create mode 100644 src/crypto/rsa/nat_test.go + +diff --git a/src/crypto/rsa/example_test.go b/src/crypto/rsa/example_test.go +index ce5c2d9..52e5639 100644 +--- a/src/crypto/rsa/example_test.go ++++ b/src/crypto/rsa/example_test.go +@@ -12,7 +12,6 @@ import ( + "crypto/sha256" + "encoding/hex" + "fmt" +- "io" + "os" + ) + +@@ -36,21 +35,17 @@ import ( + // a buffer that contains a random key. Thus, if the RSA result isn't + // well-formed, the implementation uses a random key in constant time. + func ExampleDecryptPKCS1v15SessionKey() { +- // crypto/rand.Reader is a good source of entropy for blinding the RSA +- // operation. +- rng := rand.Reader +- + // The hybrid scheme should use at least a 16-byte symmetric key. Here + // we read the random key that will be used if the RSA decryption isn't + // well-formed. + key := make([]byte, 32) +- if _, err := io.ReadFull(rng, key); err != nil { ++ if _, err := rand.Read(key); err != nil { + panic("RNG failure") + } + + rsaCiphertext, _ := hex.DecodeString("aabbccddeeff") + +- if err := DecryptPKCS1v15SessionKey(rng, rsaPrivateKey, rsaCiphertext, key); err != nil { ++ if err := DecryptPKCS1v15SessionKey(nil, rsaPrivateKey, rsaCiphertext, key); err != nil { + // Any errors that result will be “public” – meaning that they + // can be determined without any secret information. (For + // instance, if the length of key is impossible given the RSA +@@ -86,10 +81,6 @@ func ExampleDecryptPKCS1v15SessionKey() { + } + + func ExampleSignPKCS1v15() { +- // crypto/rand.Reader is a good source of entropy for blinding the RSA +- // operation. +- rng := rand.Reader +- + message := []byte("message to be signed") + + // Only small messages can be signed directly; thus the hash of a +@@ -99,7 +90,7 @@ func ExampleSignPKCS1v15() { + // of writing (2016). + hashed := sha256.Sum256(message) + +- signature, err := SignPKCS1v15(rng, rsaPrivateKey, crypto.SHA256, hashed[:]) ++ signature, err := SignPKCS1v15(nil, rsaPrivateKey, crypto.SHA256, hashed[:]) + if err != nil { + fmt.Fprintf(os.Stderr, "Error from signing: %s\n", err) + return +@@ -151,11 +142,7 @@ func ExampleDecryptOAEP() { + ciphertext, _ := hex.DecodeString("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") + label := []byte("orders") + +- // crypto/rand.Reader is a good source of entropy for blinding the RSA +- // operation. +- rng := rand.Reader +- +- plaintext, err := DecryptOAEP(sha256.New(), rng, test2048Key, ciphertext, label) ++ plaintext, err := DecryptOAEP(sha256.New(), nil, test2048Key, ciphertext, label) + if err != nil { + fmt.Fprintf(os.Stderr, "Error from decryption: %s\n", err) + return +diff --git a/src/crypto/rsa/nat.go b/src/crypto/rsa/nat.go +new file mode 100644 +index 0000000..da521c2 +--- /dev/null ++++ b/src/crypto/rsa/nat.go +@@ -0,0 +1,626 @@ ++// Copyright 2021 The Go Authors. All rights reserved. ++// Use of this source code is governed by a BSD-style ++// license that can be found in the LICENSE file. ++ ++package rsa ++ ++import ( ++ "math/big" ++ "math/bits" ++) ++ ++const ( ++ // _W is the number of bits we use for our limbs. ++ _W = bits.UintSize - 1 ++ // _MASK selects _W bits from a full machine word. ++ _MASK = (1 << _W) - 1 ++) ++ ++// choice represents a constant-time boolean. The value of choice is always ++// either 1 or 0. We use an int instead of bool in order to make decisions in ++// constant time by turning it into a mask. ++type choice uint ++ ++func not(c choice) choice { return 1 ^ c } ++ ++const yes = choice(1) ++const no = choice(0) ++ ++// ctSelect returns x if on == 1, and y if on == 0. The execution time of this ++// function does not depend on its inputs. If on is any value besides 1 or 0, ++// the result is undefined. ++func ctSelect(on choice, x, y uint) uint { ++ // When on == 1, mask is 0b111..., otherwise mask is 0b000... ++ mask := -uint(on) ++ // When mask is all zeros, we just have y, otherwise, y cancels with itself. ++ return y ^ (mask & (y ^ x)) ++} ++ ++// ctEq returns 1 if x == y, and 0 otherwise. The execution time of this ++// function does not depend on its inputs. ++func ctEq(x, y uint) choice { ++ // If x != y, then either x - y or y - x will generate a carry. ++ _, c1 := bits.Sub(x, y, 0) ++ _, c2 := bits.Sub(y, x, 0) ++ return not(choice(c1 | c2)) ++} ++ ++// ctGeq returns 1 if x >= y, and 0 otherwise. The execution time of this ++// function does not depend on its inputs. ++func ctGeq(x, y uint) choice { ++ // If x < y, then x - y generates a carry. ++ _, carry := bits.Sub(x, y, 0) ++ return not(choice(carry)) ++} ++ ++// nat represents an arbitrary natural number ++// ++// Each nat has an announced length, which is the number of limbs it has stored. ++// Operations on this number are allowed to leak this length, but will not leak ++// any information about the values contained in those limbs. ++type nat struct { ++ // limbs is a little-endian representation in base 2^W with ++ // W = bits.UintSize - 1. The top bit is always unset between operations. ++ // ++ // The top bit is left unset to optimize Montgomery multiplication, in the ++ // inner loop of exponentiation. Using fully saturated limbs would leave us ++ // working with 129-bit numbers on 64-bit platforms, wasting a lot of space, ++ // and thus time. ++ limbs []uint ++} ++ ++// expand expands x to n limbs, leaving its value unchanged. ++func (x *nat) expand(n int) *nat { ++ for len(x.limbs) > n { ++ if x.limbs[len(x.limbs)-1] != 0 { ++ panic("rsa: internal error: shrinking nat") ++ } ++ x.limbs = x.limbs[:len(x.limbs)-1] ++ } ++ if cap(x.limbs) < n { ++ newLimbs := make([]uint, n) ++ copy(newLimbs, x.limbs) ++ x.limbs = newLimbs ++ return x ++ } ++ extraLimbs := x.limbs[len(x.limbs):n] ++ for i := range extraLimbs { ++ extraLimbs[i] = 0 ++ } ++ x.limbs = x.limbs[:n] ++ return x ++} ++ ++// reset returns a zero nat of n limbs, reusing x's storage if n <= cap(x.limbs). ++func (x *nat) reset(n int) *nat { ++ if cap(x.limbs) < n { ++ x.limbs = make([]uint, n) ++ return x ++ } ++ for i := range x.limbs { ++ x.limbs[i] = 0 ++ } ++ x.limbs = x.limbs[:n] ++ return x ++} ++ ++// clone returns a new nat, with the same value and announced length as x. ++func (x *nat) clone() *nat { ++ out := &nat{make([]uint, len(x.limbs))} ++ copy(out.limbs, x.limbs) ++ return out ++} ++ ++// natFromBig creates a new natural number from a big.Int. ++// ++// The announced length of the resulting nat is based on the actual bit size of ++// the input, ignoring leading zeroes. ++func natFromBig(x *big.Int) *nat { ++ xLimbs := x.Bits() ++ bitSize := bigBitLen(x) ++ requiredLimbs := (bitSize + _W - 1) / _W ++ ++ out := &nat{make([]uint, requiredLimbs)} ++ outI := 0 ++ shift := 0 ++ for i := range xLimbs { ++ xi := uint(xLimbs[i]) ++ out.limbs[outI] |= (xi << shift) & _MASK ++ outI++ ++ if outI == requiredLimbs { ++ return out ++ } ++ out.limbs[outI] = xi >> (_W - shift) ++ shift++ // this assumes bits.UintSize - _W = 1 ++ if shift == _W { ++ shift = 0 ++ outI++ ++ } ++ } ++ return out ++} ++ ++// fillBytes sets bytes to x as a zero-extended big-endian byte slice. ++// ++// If bytes is not long enough to contain the number or at least len(x.limbs)-1 ++// limbs, or has zero length, fillBytes will panic. ++func (x *nat) fillBytes(bytes []byte) []byte { ++ if len(bytes) == 0 { ++ panic("nat: fillBytes invoked with too small buffer") ++ } ++ for i := range bytes { ++ bytes[i] = 0 ++ } ++ shift := 0 ++ outI := len(bytes) - 1 ++ for i, limb := range x.limbs { ++ remainingBits := _W ++ for remainingBits >= 8 { ++ bytes[outI] |= byte(limb) << shift ++ consumed := 8 - shift ++ limb >>= consumed ++ remainingBits -= consumed ++ shift = 0 ++ outI-- ++ if outI < 0 { ++ if limb != 0 || i < len(x.limbs)-1 { ++ panic("nat: fillBytes invoked with too small buffer") ++ } ++ return bytes ++ } ++ } ++ bytes[outI] = byte(limb) ++ shift = remainingBits ++ } ++ return bytes ++} ++ ++// natFromBytes converts a slice of big-endian bytes into a nat. ++// ++// The announced length of the output depends on the length of bytes. Unlike ++// big.Int, creating a nat will not remove leading zeros. ++func natFromBytes(bytes []byte) *nat { ++ bitSize := len(bytes) * 8 ++ requiredLimbs := (bitSize + _W - 1) / _W ++ ++ out := &nat{make([]uint, requiredLimbs)} ++ outI := 0 ++ shift := 0 ++ for i := len(bytes) - 1; i >= 0; i-- { ++ bi := bytes[i] ++ out.limbs[outI] |= uint(bi) << shift ++ shift += 8 ++ if shift >= _W { ++ shift -= _W ++ out.limbs[outI] &= _MASK ++ outI++ ++ if shift > 0 { ++ out.limbs[outI] = uint(bi) >> (8 - shift) ++ } ++ } ++ } ++ return out ++} ++ ++// cmpEq returns 1 if x == y, and 0 otherwise. ++// ++// Both operands must have the same announced length. ++func (x *nat) cmpEq(y *nat) choice { ++ // Eliminate bounds checks in the loop. ++ size := len(x.limbs) ++ xLimbs := x.limbs[:size] ++ yLimbs := y.limbs[:size] ++ ++ equal := yes ++ for i := 0; i < size; i++ { ++ equal &= ctEq(xLimbs[i], yLimbs[i]) ++ } ++ return equal ++} ++ ++// cmpGeq returns 1 if x >= y, and 0 otherwise. ++// ++// Both operands must have the same announced length. ++func (x *nat) cmpGeq(y *nat) choice { ++ // Eliminate bounds checks in the loop. ++ size := len(x.limbs) ++ xLimbs := x.limbs[:size] ++ yLimbs := y.limbs[:size] ++ ++ var c uint ++ for i := 0; i < size; i++ { ++ c = (xLimbs[i] - yLimbs[i] - c) >> _W ++ } ++ // If there was a carry, then subtracting y underflowed, so ++ // x is not greater than or equal to y. ++ return not(choice(c)) ++} ++ ++// assign sets x <- y if on == 1, and does nothing otherwise. ++// ++// Both operands must have the same announced length. ++func (x *nat) assign(on choice, y *nat) *nat { ++ // Eliminate bounds checks in the loop. ++ size := len(x.limbs) ++ xLimbs := x.limbs[:size] ++ yLimbs := y.limbs[:size] ++ ++ for i := 0; i < size; i++ { ++ xLimbs[i] = ctSelect(on, yLimbs[i], xLimbs[i]) ++ } ++ return x ++} ++ ++// add computes x += y if on == 1, and does nothing otherwise. It returns the ++// carry of the addition regardless of on. ++// ++// Both operands must have the same announced length. ++func (x *nat) add(on choice, y *nat) (c uint) { ++ // Eliminate bounds checks in the loop. ++ size := len(x.limbs) ++ xLimbs := x.limbs[:size] ++ yLimbs := y.limbs[:size] ++ ++ for i := 0; i < size; i++ { ++ res := xLimbs[i] + yLimbs[i] + c ++ xLimbs[i] = ctSelect(on, res&_MASK, xLimbs[i]) ++ c = res >> _W ++ } ++ return ++} ++ ++// sub computes x -= y if on == 1, and does nothing otherwise. It returns the ++// borrow of the subtraction regardless of on. ++// ++// Both operands must have the same announced length. ++func (x *nat) sub(on choice, y *nat) (c uint) { ++ // Eliminate bounds checks in the loop. ++ size := len(x.limbs) ++ xLimbs := x.limbs[:size] ++ yLimbs := y.limbs[:size] ++ ++ for i := 0; i < size; i++ { ++ res := xLimbs[i] - yLimbs[i] - c ++ xLimbs[i] = ctSelect(on, res&_MASK, xLimbs[i]) ++ c = res >> _W ++ } ++ return ++} ++ ++// modulus is used for modular arithmetic, precomputing relevant constants. ++// ++// Moduli are assumed to be odd numbers. Moduli can also leak the exact ++// number of bits needed to store their value, and are stored without padding. ++// ++// Their actual value is still kept secret. ++type modulus struct { ++ // The underlying natural number for this modulus. ++ // ++ // This will be stored without any padding, and shouldn't alias with any ++ // other natural number being used. ++ nat *nat ++ leading int // number of leading zeros in the modulus ++ m0inv uint // -nat.limbs[0]⁻¹ mod _W ++} ++ ++// minusInverseModW computes -x⁻¹ mod _W with x odd. ++// ++// This operation is used to precompute a constant involved in Montgomery ++// multiplication. ++func minusInverseModW(x uint) uint { ++ // Every iteration of this loop doubles the least-significant bits of ++ // correct inverse in y. The first three bits are already correct (1⁻¹ = 1, ++ // 3⁻¹ = 3, 5⁻¹ = 5, and 7⁻¹ = 7 mod 8), so doubling five times is enough ++ // for 61 bits (and wastes only one iteration for 31 bits). ++ // ++ // See https://crypto.stackexchange.com/a/47496. ++ y := x ++ for i := 0; i < 5; i++ { ++ y = y * (2 - x*y) ++ } ++ return (1 << _W) - (y & _MASK) ++} ++ ++// modulusFromNat creates a new modulus from a nat. ++// ++// The nat should be odd, nonzero, and the number of significant bits in the ++// number should be leakable. The nat shouldn't be reused. ++func modulusFromNat(nat *nat) *modulus { ++ m := &modulus{} ++ m.nat = nat ++ size := len(m.nat.limbs) ++ for m.nat.limbs[size-1] == 0 { ++ size-- ++ } ++ m.nat.limbs = m.nat.limbs[:size] ++ m.leading = _W - bitLen(m.nat.limbs[size-1]) ++ m.m0inv = minusInverseModW(m.nat.limbs[0]) ++ return m ++} ++ ++// bitLen is a version of bits.Len that only leaks the bit length of n, but not ++// its value. bits.Len and bits.LeadingZeros use a lookup table for the ++// low-order bits on some architectures. ++func bitLen(n uint) int { ++ var len int ++ // We assume, here and elsewhere, that comparison to zero is constant time ++ // with respect to different non-zero values. ++ for n != 0 { ++ len++ ++ n >>= 1 ++ } ++ return len ++} ++ ++// bigBitLen is a version of big.Int.BitLen that only leaks the bit length of x, ++// but not its value. big.Int.BitLen uses bits.Len. ++func bigBitLen(x *big.Int) int { ++ xLimbs := x.Bits() ++ fullLimbs := len(xLimbs) - 1 ++ topLimb := uint(xLimbs[len(xLimbs)-1]) ++ return fullLimbs*bits.UintSize + bitLen(topLimb) ++} ++ ++// modulusSize returns the size of m in bytes. ++func modulusSize(m *modulus) int { ++ bits := len(m.nat.limbs)*_W - int(m.leading) ++ return (bits + 7) / 8 ++} ++ ++// shiftIn calculates x = x << _W + y mod m. ++// ++// This assumes that x is already reduced mod m, and that y < 2^_W. ++func (x *nat) shiftIn(y uint, m *modulus) *nat { ++ d := new(nat).resetFor(m) ++ ++ // Eliminate bounds checks in the loop. ++ size := len(m.nat.limbs) ++ xLimbs := x.limbs[:size] ++ dLimbs := d.limbs[:size] ++ mLimbs := m.nat.limbs[:size] ++ ++ // Each iteration of this loop computes x = 2x + b mod m, where b is a bit ++ // from y. Effectively, it left-shifts x and adds y one bit at a time, ++ // reducing it every time. ++ // ++ // To do the reduction, each iteration computes both 2x + b and 2x + b - m. ++ // The next iteration (and finally the return line) will use either result ++ // based on whether the subtraction underflowed. ++ needSubtraction := no ++ for i := _W - 1; i >= 0; i-- { ++ carry := (y >> i) & 1 ++ var borrow uint ++ for i := 0; i < size; i++ { ++ l := ctSelect(needSubtraction, dLimbs[i], xLimbs[i]) ++ ++ res := l<<1 + carry ++ xLimbs[i] = res & _MASK ++ carry = res >> _W ++ ++ res = xLimbs[i] - mLimbs[i] - borrow ++ dLimbs[i] = res & _MASK ++ borrow = res >> _W ++ } ++ // See modAdd for how carry (aka overflow), borrow (aka underflow), and ++ // needSubtraction relate. ++ needSubtraction = ctEq(carry, borrow) ++ } ++ return x.assign(needSubtraction, d) ++} ++ ++// mod calculates out = x mod m. ++// ++// This works regardless how large the value of x is. ++// ++// The output will be resized to the size of m and overwritten. ++func (out *nat) mod(x *nat, m *modulus) *nat { ++ out.resetFor(m) ++ // Working our way from the most significant to the least significant limb, ++ // we can insert each limb at the least significant position, shifting all ++ // previous limbs left by _W. This way each limb will get shifted by the ++ // correct number of bits. We can insert at least N - 1 limbs without ++ // overflowing m. After that, we need to reduce every time we shift. ++ i := len(x.limbs) - 1 ++ // For the first N - 1 limbs we can skip the actual shifting and position ++ // them at the shifted position, which starts at min(N - 2, i). ++ start := len(m.nat.limbs) - 2 ++ if i < start { ++ start = i ++ } ++ for j := start; j >= 0; j-- { ++ out.limbs[j] = x.limbs[i] ++ i-- ++ } ++ // We shift in the remaining limbs, reducing modulo m each time. ++ for i >= 0 { ++ out.shiftIn(x.limbs[i], m) ++ i-- ++ } ++ return out ++} ++ ++// expandFor ensures out has the right size to work with operations modulo m. ++// ++// This assumes that out has as many or fewer limbs than m, or that the extra ++// limbs are all zero (which may happen when decoding a value that has leading ++// zeroes in its bytes representation that spill over the limb threshold). ++func (out *nat) expandFor(m *modulus) *nat { ++ return out.expand(len(m.nat.limbs)) ++} ++ ++// resetFor ensures out has the right size to work with operations modulo m. ++// ++// out is zeroed and may start at any size. ++func (out *nat) resetFor(m *modulus) *nat { ++ return out.reset(len(m.nat.limbs)) ++} ++ ++// modSub computes x = x - y mod m. ++// ++// The length of both operands must be the same as the modulus. Both operands ++// must already be reduced modulo m. ++func (x *nat) modSub(y *nat, m *modulus) *nat { ++ underflow := x.sub(yes, y) ++ // If the subtraction underflowed, add m. ++ x.add(choice(underflow), m.nat) ++ return x ++} ++ ++// modAdd computes x = x + y mod m. ++// ++// The length of both operands must be the same as the modulus. Both operands ++// must already be reduced modulo m. ++func (x *nat) modAdd(y *nat, m *modulus) *nat { ++ overflow := x.add(yes, y) ++ underflow := not(x.cmpGeq(m.nat)) // x < m ++ ++ // Three cases are possible: ++ // ++ // - overflow = 0, underflow = 0 ++ // ++ // In this case, addition fits in our limbs, but we can still subtract away ++ // m without an underflow, so we need to perform the subtraction to reduce ++ // our result. ++ // ++ // - overflow = 0, underflow = 1 ++ // ++ // The addition fits in our limbs, but we can't subtract m without ++ // underflowing. The result is already reduced. ++ // ++ // - overflow = 1, underflow = 1 ++ // ++ // The addition does not fit in our limbs, and the subtraction's borrow ++ // would cancel out with the addition's carry. We need to subtract m to ++ // reduce our result. ++ // ++ // The overflow = 1, underflow = 0 case is not possible, because y is at ++ // most m - 1, and if adding m - 1 overflows, then subtracting m must ++ // necessarily underflow. ++ needSubtraction := ctEq(overflow, uint(underflow)) ++ ++ x.sub(needSubtraction, m.nat) ++ return x ++} ++ ++// montgomeryRepresentation calculates x = x * R mod m, with R = 2^(_W * n) and ++// n = len(m.nat.limbs). ++// ++// Faster Montgomery multiplication replaces standard modular multiplication for ++// numbers in this representation. ++// ++// This assumes that x is already reduced mod m. ++func (x *nat) montgomeryRepresentation(m *modulus) *nat { ++ for i := 0; i < len(m.nat.limbs); i++ { ++ x.shiftIn(0, m) // x = x * 2^_W mod m ++ } ++ return x ++} ++ ++// montgomeryMul calculates d = a * b / R mod m, with R = 2^(_W * n) and ++// n = len(m.nat.limbs), using the Montgomery Multiplication technique. ++// ++// All inputs should be the same length, not aliasing d, and already ++// reduced modulo m. d will be resized to the size of m and overwritten. ++func (d *nat) montgomeryMul(a *nat, b *nat, m *modulus) *nat { ++ // See https://bearssl.org/bigint.html#montgomery-reduction-and-multiplication ++ // for a description of the algorithm. ++ ++ // Eliminate bounds checks in the loop. ++ size := len(m.nat.limbs) ++ aLimbs := a.limbs[:size] ++ bLimbs := b.limbs[:size] ++ dLimbs := d.resetFor(m).limbs[:size] ++ mLimbs := m.nat.limbs[:size] ++ ++ var overflow uint ++ for i := 0; i < size; i++ { ++ f := ((dLimbs[0] + aLimbs[i]*bLimbs[0]) * m.m0inv) & _MASK ++ carry := uint(0) ++ for j := 0; j < size; j++ { ++ // z = d[j] + a[i] * b[j] + f * m[j] + carry <= 2^(2W+1) - 2^(W+1) + 2^W ++ hi, lo := bits.Mul(aLimbs[i], bLimbs[j]) ++ z_lo, c := bits.Add(dLimbs[j], lo, 0) ++ z_hi, _ := bits.Add(0, hi, c) ++ hi, lo = bits.Mul(f, mLimbs[j]) ++ z_lo, c = bits.Add(z_lo, lo, 0) ++ z_hi, _ = bits.Add(z_hi, hi, c) ++ z_lo, c = bits.Add(z_lo, carry, 0) ++ z_hi, _ = bits.Add(z_hi, 0, c) ++ if j > 0 { ++ dLimbs[j-1] = z_lo & _MASK ++ } ++ carry = z_hi<<1 | z_lo>>_W // carry <= 2^(W+1) - 2 ++ } ++ z := overflow + carry // z <= 2^(W+1) - 1 ++ dLimbs[size-1] = z & _MASK ++ overflow = z >> _W // overflow <= 1 ++ } ++ // See modAdd for how overflow, underflow, and needSubtraction relate. ++ underflow := not(d.cmpGeq(m.nat)) // d < m ++ needSubtraction := ctEq(overflow, uint(underflow)) ++ d.sub(needSubtraction, m.nat) ++ ++ return d ++} ++ ++// modMul calculates x *= y mod m. ++// ++// x and y must already be reduced modulo m, they must share its announced ++// length, and they may not alias. ++func (x *nat) modMul(y *nat, m *modulus) *nat { ++ // A Montgomery multiplication by a value out of the Montgomery domain ++ // takes the result out of Montgomery representation. ++ xR := x.clone().montgomeryRepresentation(m) // xR = x * R mod m ++ return x.montgomeryMul(xR, y, m) // x = xR * y / R mod m ++} ++ ++// exp calculates out = x^e mod m. ++// ++// The exponent e is represented in big-endian order. The output will be resized ++// to the size of m and overwritten. x must already be reduced modulo m. ++func (out *nat) exp(x *nat, e []byte, m *modulus) *nat { ++ // We use a 4 bit window. For our RSA workload, 4 bit windows are faster ++ // than 2 bit windows, but use an extra 12 nats worth of scratch space. ++ // Using bit sizes that don't divide 8 are more complex to implement. ++ table := make([]*nat, (1<<4)-1) // table[i] = x ^ (i+1) ++ table[0] = x.clone().montgomeryRepresentation(m) ++ for i := 1; i < len(table); i++ { ++ table[i] = new(nat).expandFor(m) ++ table[i].montgomeryMul(table[i-1], table[0], m) ++ } ++ ++ out.resetFor(m) ++ out.limbs[0] = 1 ++ out.montgomeryRepresentation(m) ++ t0 := new(nat).expandFor(m) ++ t1 := new(nat).expandFor(m) ++ for _, b := range e { ++ for _, j := range []int{4, 0} { ++ // Square four times. ++ t1.montgomeryMul(out, out, m) ++ out.montgomeryMul(t1, t1, m) ++ t1.montgomeryMul(out, out, m) ++ out.montgomeryMul(t1, t1, m) ++ ++ // Select x^k in constant time from the table. ++ k := uint((b >> j) & 0b1111) ++ for i := range table { ++ t0.assign(ctEq(k, uint(i+1)), table[i]) ++ } ++ ++ // Multiply by x^k, discarding the result if k = 0. ++ t1.montgomeryMul(out, t0, m) ++ out.assign(not(ctEq(k, 0)), t1) ++ } ++ } ++ ++ // By Montgomery multiplying with 1 not in Montgomery representation, we ++ // convert out back from Montgomery representation, because it works out to ++ // dividing by R. ++ t0.assign(yes, out) ++ t1.resetFor(m) ++ t1.limbs[0] = 1 ++ out.montgomeryMul(t0, t1, m) ++ ++ return out ++} +diff --git a/src/crypto/rsa/nat_test.go b/src/crypto/rsa/nat_test.go +new file mode 100644 +index 0000000..3e6eb10 +--- /dev/null ++++ b/src/crypto/rsa/nat_test.go +@@ -0,0 +1,384 @@ ++// Copyright 2021 The Go Authors. All rights reserved. ++// Use of this source code is governed by a BSD-style ++// license that can be found in the LICENSE file. ++ ++package rsa ++ ++import ( ++ "bytes" ++ "math/big" ++ "math/bits" ++ "math/rand" ++ "reflect" ++ "testing" ++ "testing/quick" ++) ++ ++// Generate generates an even nat. It's used by testing/quick to produce random ++// *nat values for quick.Check invocations. ++func (*nat) Generate(r *rand.Rand, size int) reflect.Value { ++ limbs := make([]uint, size) ++ for i := 0; i < size; i++ { ++ limbs[i] = uint(r.Uint64()) & ((1 << _W) - 2) ++ } ++ return reflect.ValueOf(&nat{limbs}) ++} ++ ++func testModAddCommutative(a *nat, b *nat) bool { ++ mLimbs := make([]uint, len(a.limbs)) ++ for i := 0; i < len(mLimbs); i++ { ++ mLimbs[i] = _MASK ++ } ++ m := modulusFromNat(&nat{mLimbs}) ++ aPlusB := a.clone() ++ aPlusB.modAdd(b, m) ++ bPlusA := b.clone() ++ bPlusA.modAdd(a, m) ++ return aPlusB.cmpEq(bPlusA) == 1 ++} ++ ++func TestModAddCommutative(t *testing.T) { ++ err := quick.Check(testModAddCommutative, &quick.Config{}) ++ if err != nil { ++ t.Error(err) ++ } ++} ++ ++func testModSubThenAddIdentity(a *nat, b *nat) bool { ++ mLimbs := make([]uint, len(a.limbs)) ++ for i := 0; i < len(mLimbs); i++ { ++ mLimbs[i] = _MASK ++ } ++ m := modulusFromNat(&nat{mLimbs}) ++ original := a.clone() ++ a.modSub(b, m) ++ a.modAdd(b, m) ++ return a.cmpEq(original) == 1 ++} ++ ++func TestModSubThenAddIdentity(t *testing.T) { ++ err := quick.Check(testModSubThenAddIdentity, &quick.Config{}) ++ if err != nil { ++ t.Error(err) ++ } ++} ++ ++func testMontgomeryRoundtrip(a *nat) bool { ++ one := &nat{make([]uint, len(a.limbs))} ++ one.limbs[0] = 1 ++ aPlusOne := a.clone() ++ aPlusOne.add(1, one) ++ m := modulusFromNat(aPlusOne) ++ monty := a.clone() ++ monty.montgomeryRepresentation(m) ++ aAgain := monty.clone() ++ aAgain.montgomeryMul(monty, one, m) ++ return a.cmpEq(aAgain) == 1 ++} ++ ++func TestMontgomeryRoundtrip(t *testing.T) { ++ err := quick.Check(testMontgomeryRoundtrip, &quick.Config{}) ++ if err != nil { ++ t.Error(err) ++ } ++} ++ ++func TestFromBig(t *testing.T) { ++ expected := []byte{0x01, 0xff, 0xff, 0xff, 0xff, 0xff, 0xff, 0xff, 0xff, 0xff, 0xff, 0xff, 0xff, 0xff, 0xff, 0xff, 0xff} ++ theBig := new(big.Int).SetBytes(expected) ++ actual := natFromBig(theBig).fillBytes(make([]byte, len(expected))) ++ if !bytes.Equal(actual, expected) { ++ t.Errorf("%+x != %+x", actual, expected) ++ } ++} ++ ++func TestFillBytes(t *testing.T) { ++ xBytes := []byte{0xAA, 0xFF, 0x22, 0x33, 0x44, 0x55, 0x66, 0x77, 0x88} ++ x := natFromBytes(xBytes) ++ for l := 20; l >= len(xBytes); l-- { ++ buf := make([]byte, l) ++ rand.Read(buf) ++ actual := x.fillBytes(buf) ++ expected := make([]byte, l) ++ copy(expected[l-len(xBytes):], xBytes) ++ if !bytes.Equal(actual, expected) { ++ t.Errorf("%d: %+v != %+v", l, actual, expected) ++ } ++ } ++ for l := len(xBytes) - 1; l >= 0; l-- { ++ (func() { ++ defer func() { ++ if recover() == nil { ++ t.Errorf("%d: expected panic", l) ++ } ++ }() ++ x.fillBytes(make([]byte, l)) ++ })() ++ } ++} ++ ++func TestFromBytes(t *testing.T) { ++ f := func(xBytes []byte) bool { ++ if len(xBytes) == 0 { ++ return true ++ } ++ actual := natFromBytes(xBytes).fillBytes(make([]byte, len(xBytes))) ++ if !bytes.Equal(actual, xBytes) { ++ t.Errorf("%+x != %+x", actual, xBytes) ++ return false ++ } ++ return true ++ } ++ ++ err := quick.Check(f, &quick.Config{}) ++ if err != nil { ++ t.Error(err) ++ } ++ ++ f([]byte{0xFF, 0x22, 0x33, 0x44, 0x55, 0x66, 0x77, 0x88}) ++ f(bytes.Repeat([]byte{0xFF}, _W)) ++} ++ ++func TestShiftIn(t *testing.T) { ++ if bits.UintSize != 64 { ++ t.Skip("examples are only valid in 64 bit") ++ } ++ examples := []struct { ++ m, x, expected []byte ++ y uint64 ++ }{{ ++ m: []byte{13}, ++ x: []byte{0}, ++ y: 0x7FFF_FFFF_FFFF_FFFF, ++ expected: []byte{7}, ++ }, { ++ m: []byte{13}, ++ x: []byte{7}, ++ y: 0x7FFF_FFFF_FFFF_FFFF, ++ expected: []byte{11}, ++ }, { ++ m: []byte{0x06, 0x80, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x0d}, ++ x: make([]byte, 9), ++ y: 0x7FFF_FFFF_FFFF_FFFF, ++ expected: []byte{0x00, 0x7f, 0xff, 0xff, 0xff, 0xff, 0xff, 0xff, 0xff}, ++ }, { ++ m: []byte{0x06, 0x80, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x0d}, ++ x: []byte{0x00, 0x7f, 0xff, 0xff, 0xff, 0xff, 0xff, 0xff, 0xff}, ++ y: 0, ++ expected: []byte{0x03, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x08}, ++ }} ++ ++ for i, tt := range examples { ++ m := modulusFromNat(natFromBytes(tt.m)) ++ got := natFromBytes(tt.x).expandFor(m).shiftIn(uint(tt.y), m) ++ if got.cmpEq(natFromBytes(tt.expected).expandFor(m)) != 1 { ++ t.Errorf("%d: got %x, expected %x", i, got, tt.expected) ++ } ++ } ++} ++ ++func TestModulusAndNatSizes(t *testing.T) { ++ // These are 126 bit (2 * _W on 64-bit architectures) values, serialized as ++ // 128 bits worth of bytes. If leading zeroes are stripped, they fit in two ++ // limbs, if they are not, they fit in three. This can be a problem because ++ // modulus strips leading zeroes and nat does not. ++ m := modulusFromNat(natFromBytes([]byte{ ++ 0x3f, 0xff, 0xff, 0xff, 0xff, 0xff, 0xff, 0xff, ++ 0xff, 0xff, 0xff, 0xff, 0xff, 0xff, 0xff, 0xff})) ++ x := natFromBytes([]byte{ ++ 0x3f, 0xff, 0xff, 0xff, 0xff, 0xff, 0xff, 0xff, ++ 0xff, 0xff, 0xff, 0xff, 0xff, 0xff, 0xff, 0xfe}) ++ x.expandFor(m) // must not panic for shrinking ++} ++ ++func TestExpand(t *testing.T) { ++ sliced := []uint{1, 2, 3, 4} ++ examples := []struct { ++ in []uint ++ n int ++ out []uint ++ }{{ ++ []uint{1, 2}, ++ 4, ++ []uint{1, 2, 0, 0}, ++ }, { ++ sliced[:2], ++ 4, ++ []uint{1, 2, 0, 0}, ++ }, { ++ []uint{1, 2}, ++ 2, ++ []uint{1, 2}, ++ }, { ++ []uint{1, 2, 0}, ++ 2, ++ []uint{1, 2}, ++ }} ++ ++ for i, tt := range examples { ++ got := (&nat{tt.in}).expand(tt.n) ++ if len(got.limbs) != len(tt.out) || got.cmpEq(&nat{tt.out}) != 1 { ++ t.Errorf("%d: got %x, expected %x", i, got, tt.out) ++ } ++ } ++} ++ ++func TestMod(t *testing.T) { ++ m := modulusFromNat(natFromBytes([]byte{0x06, 0x80, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x0d})) ++ x := natFromBytes([]byte{0x40, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x80, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x01}) ++ out := new(nat) ++ out.mod(x, m) ++ expected := natFromBytes([]byte{0x04, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x09}) ++ if out.cmpEq(expected) != 1 { ++ t.Errorf("%+v != %+v", out, expected) ++ } ++} ++ ++func TestModSub(t *testing.T) { ++ m := modulusFromNat(&nat{[]uint{13}}) ++ x := &nat{[]uint{6}} ++ y := &nat{[]uint{7}} ++ x.modSub(y, m) ++ expected := &nat{[]uint{12}} ++ if x.cmpEq(expected) != 1 { ++ t.Errorf("%+v != %+v", x, expected) ++ } ++ x.modSub(y, m) ++ expected = &nat{[]uint{5}} ++ if x.cmpEq(expected) != 1 { ++ t.Errorf("%+v != %+v", x, expected) ++ } ++} ++ ++func TestModAdd(t *testing.T) { ++ m := modulusFromNat(&nat{[]uint{13}}) ++ x := &nat{[]uint{6}} ++ y := &nat{[]uint{7}} ++ x.modAdd(y, m) ++ expected := &nat{[]uint{0}} ++ if x.cmpEq(expected) != 1 { ++ t.Errorf("%+v != %+v", x, expected) ++ } ++ x.modAdd(y, m) ++ expected = &nat{[]uint{7}} ++ if x.cmpEq(expected) != 1 { ++ t.Errorf("%+v != %+v", x, expected) ++ } ++} ++ ++func TestExp(t *testing.T) { ++ m := modulusFromNat(&nat{[]uint{13}}) ++ x := &nat{[]uint{3}} ++ out := &nat{[]uint{0}} ++ out.exp(x, []byte{12}, m) ++ expected := &nat{[]uint{1}} ++ if out.cmpEq(expected) != 1 { ++ t.Errorf("%+v != %+v", out, expected) ++ } ++} ++ ++func makeBenchmarkModulus() *modulus { ++ m := make([]uint, 32) ++ for i := 0; i < 32; i++ { ++ m[i] = _MASK ++ } ++ return modulusFromNat(&nat{limbs: m}) ++} ++ ++func makeBenchmarkValue() *nat { ++ x := make([]uint, 32) ++ for i := 0; i < 32; i++ { ++ x[i] = _MASK - 1 ++ } ++ return &nat{limbs: x} ++} ++ ++func makeBenchmarkExponent() []byte { ++ e := make([]byte, 256) ++ for i := 0; i < 32; i++ { ++ e[i] = 0xFF ++ } ++ return e ++} ++ ++func BenchmarkModAdd(b *testing.B) { ++ x := makeBenchmarkValue() ++ y := makeBenchmarkValue() ++ m := makeBenchmarkModulus() ++ ++ b.ResetTimer() ++ for i := 0; i < b.N; i++ { ++ x.modAdd(y, m) ++ } ++} ++ ++func BenchmarkModSub(b *testing.B) { ++ x := makeBenchmarkValue() ++ y := makeBenchmarkValue() ++ m := makeBenchmarkModulus() ++ ++ b.ResetTimer() ++ for i := 0; i < b.N; i++ { ++ x.modSub(y, m) ++ } ++} ++ ++func BenchmarkMontgomeryRepr(b *testing.B) { ++ x := makeBenchmarkValue() ++ m := makeBenchmarkModulus() ++ ++ b.ResetTimer() ++ for i := 0; i < b.N; i++ { ++ x.montgomeryRepresentation(m) ++ } ++} ++ ++func BenchmarkMontgomeryMul(b *testing.B) { ++ x := makeBenchmarkValue() ++ y := makeBenchmarkValue() ++ out := makeBenchmarkValue() ++ m := makeBenchmarkModulus() ++ ++ b.ResetTimer() ++ for i := 0; i < b.N; i++ { ++ out.montgomeryMul(x, y, m) ++ } ++} ++ ++func BenchmarkModMul(b *testing.B) { ++ x := makeBenchmarkValue() ++ y := makeBenchmarkValue() ++ m := makeBenchmarkModulus() ++ ++ b.ResetTimer() ++ for i := 0; i < b.N; i++ { ++ x.modMul(y, m) ++ } ++} ++ ++func BenchmarkExpBig(b *testing.B) { ++ out := new(big.Int) ++ exponentBytes := makeBenchmarkExponent() ++ x := new(big.Int).SetBytes(exponentBytes) ++ e := new(big.Int).SetBytes(exponentBytes) ++ n := new(big.Int).SetBytes(exponentBytes) ++ one := new(big.Int).SetUint64(1) ++ n.Add(n, one) ++ ++ b.ResetTimer() ++ for i := 0; i < b.N; i++ { ++ out.Exp(x, e, n) ++ } ++} ++ ++func BenchmarkExp(b *testing.B) { ++ x := makeBenchmarkValue() ++ e := makeBenchmarkExponent() ++ out := makeBenchmarkValue() ++ m := makeBenchmarkModulus() ++ ++ b.ResetTimer() ++ for i := 0; i < b.N; i++ { ++ out.exp(x, e, m) ++ } ++} +diff --git a/src/crypto/rsa/pkcs1v15.go b/src/crypto/rsa/pkcs1v15.go +index 0cbd6d0..90233bb 100644 +--- a/src/crypto/rsa/pkcs1v15.go ++++ b/src/crypto/rsa/pkcs1v15.go +@@ -9,7 +9,6 @@ import ( + "crypto/subtle" + "errors" + "io" +- "math/big" + + "crypto/internal/randutil" + ) +@@ -58,14 +57,11 @@ func EncryptPKCS1v15(rand io.Reader, pub *PublicKey, msg []byte) ([]byte, error) + em[len(em)-len(msg)-1] = 0 + copy(mm, msg) + +- m := new(big.Int).SetBytes(em) +- c := encrypt(new(big.Int), pub, m) +- +- return c.FillBytes(em), nil ++ return encrypt(pub, em), nil + } + + // DecryptPKCS1v15 decrypts a plaintext using RSA and the padding scheme from PKCS #1 v1.5. +-// If rand != nil, it uses RSA blinding to avoid timing side-channel attacks. ++// The random parameter is legacy and ignored, and it can be as nil. + // + // Note that whether this function returns an error or not discloses secret + // information. If an attacker can cause this function to run repeatedly and +@@ -76,7 +72,7 @@ func DecryptPKCS1v15(rand io.Reader, priv *PrivateKey, ciphertext []byte) ([]byt + if err := checkPub(&priv.PublicKey); err != nil { + return nil, err + } +- valid, out, index, err := decryptPKCS1v15(rand, priv, ciphertext) ++ valid, out, index, err := decryptPKCS1v15(priv, ciphertext) + if err != nil { + return nil, err + } +@@ -87,7 +83,7 @@ func DecryptPKCS1v15(rand io.Reader, priv *PrivateKey, ciphertext []byte) ([]byt + } + + // DecryptPKCS1v15SessionKey decrypts a session key using RSA and the padding scheme from PKCS #1 v1.5. +-// If rand != nil, it uses RSA blinding to avoid timing side-channel attacks. ++// The random parameter is legacy and ignored, and it can be as nil. + // It returns an error if the ciphertext is the wrong length or if the + // ciphertext is greater than the public modulus. Otherwise, no error is + // returned. If the padding is valid, the resulting plaintext message is copied +@@ -114,7 +110,7 @@ func DecryptPKCS1v15SessionKey(rand io.Reader, priv *PrivateKey, ciphertext []by + return ErrDecryption + } + +- valid, em, index, err := decryptPKCS1v15(rand, priv, ciphertext) ++ valid, em, index, err := decryptPKCS1v15(priv, ciphertext) + if err != nil { + return err + } +@@ -130,26 +126,24 @@ func DecryptPKCS1v15SessionKey(rand io.Reader, priv *PrivateKey, ciphertext []by + return nil + } + +-// decryptPKCS1v15 decrypts ciphertext using priv and blinds the operation if +-// rand is not nil. It returns one or zero in valid that indicates whether the +-// plaintext was correctly structured. In either case, the plaintext is +-// returned in em so that it may be read independently of whether it was valid +-// in order to maintain constant memory access patterns. If the plaintext was +-// valid then index contains the index of the original message in em. +-func decryptPKCS1v15(rand io.Reader, priv *PrivateKey, ciphertext []byte) (valid int, em []byte, index int, err error) { ++// decryptPKCS1v15 decrypts ciphertext using priv. It returns one or zero in ++// valid that indicates whether the plaintext was correctly structured. ++// In either case, the plaintext is returned in em so that it may be read ++// independently of whether it was valid in order to maintain constant memory ++// access patterns. If the plaintext was valid then index contains the index of ++// the original message in em, to allow constant time padding removal. ++func decryptPKCS1v15(priv *PrivateKey, ciphertext []byte) (valid int, em []byte, index int, err error) { + k := priv.Size() + if k < 11 { + err = ErrDecryption + return + } + +- c := new(big.Int).SetBytes(ciphertext) +- m, err := decrypt(rand, priv, c) ++ em, err = decrypt(priv, ciphertext) + if err != nil { + return + } + +- em = m.FillBytes(make([]byte, k)) + firstByteIsZero := subtle.ConstantTimeByteEq(em[0], 0) + secondByteIsTwo := subtle.ConstantTimeByteEq(em[1], 2) + +@@ -221,8 +215,7 @@ var hashPrefixes = map[crypto.Hash][]byte{ + // function. If hash is zero, hashed is signed directly. This isn't + // advisable except for interoperability. + // +-// If rand is not nil then RSA blinding will be used to avoid timing +-// side-channel attacks. ++// The random parameter is legacy and ignored, and it can be as nil. + // + // This function is deterministic. Thus, if the set of possible + // messages is small, an attacker may be able to build a map from +@@ -249,13 +242,7 @@ func SignPKCS1v15(rand io.Reader, priv *PrivateKey, hash crypto.Hash, hashed []b + copy(em[k-tLen:k-hashLen], prefix) + copy(em[k-hashLen:k], hashed) + +- m := new(big.Int).SetBytes(em) +- c, err := decryptAndCheck(rand, priv, m) +- if err != nil { +- return nil, err +- } +- +- return c.FillBytes(em), nil ++ return decryptAndCheck(priv, em) + } + + // VerifyPKCS1v15 verifies an RSA PKCS #1 v1.5 signature. +@@ -282,9 +269,7 @@ func VerifyPKCS1v15(pub *PublicKey, hash crypto.Hash, hashed []byte, sig []byte) + return ErrVerification + } + +- c := new(big.Int).SetBytes(sig) +- m := encrypt(new(big.Int), pub, c) +- em := m.FillBytes(make([]byte, k)) ++ em := encrypt(pub, sig) + // EM = 0x00 || 0x01 || PS || 0x00 || T + + ok := subtle.ConstantTimeByteEq(em[0], 0) +diff --git a/src/crypto/rsa/pss.go b/src/crypto/rsa/pss.go +index 814522d..aeb6148 100644 +--- a/src/crypto/rsa/pss.go ++++ b/src/crypto/rsa/pss.go +@@ -12,7 +12,6 @@ import ( + "errors" + "hash" + "io" +- "math/big" + ) + + // Per RFC 8017, Section 9.1 +@@ -207,19 +206,26 @@ func emsaPSSVerify(mHash, em []byte, emBits, sLen int, hash hash.Hash) error { + // Note that hashed must be the result of hashing the input message using the + // given hash function. salt is a random sequence of bytes whose length will be + // later used to verify the signature. +-func signPSSWithSalt(rand io.Reader, priv *PrivateKey, hash crypto.Hash, hashed, salt []byte) ([]byte, error) { +- emBits := priv.N.BitLen() - 1 ++func signPSSWithSalt(priv *PrivateKey, hash crypto.Hash, hashed, salt []byte) ([]byte, error) { ++ emBits := bigBitLen(priv.N) - 1 + em, err := emsaPSSEncode(hashed, emBits, salt, hash.New()) + if err != nil { + return nil, err + } +- m := new(big.Int).SetBytes(em) +- c, err := decryptAndCheck(rand, priv, m) +- if err != nil { +- return nil, err ++ // RFC 8017: "Note that the octet length of EM will be one less than k if ++ // modBits - 1 is divisible by 8 and equal to k otherwise, where k is the ++ // length in octets of the RSA modulus n." ++ // ++ // This is extremely annoying, as all other encrypt and decrypt inputs are ++ // always the exact same size as the modulus. Since it only happens for ++ // weird modulus sizes, fix it by padding inefficiently. ++ if emLen, k := len(em), priv.Size(); emLen < k { ++ emNew := make([]byte, k) ++ copy(emNew[k-emLen:], em) ++ em = emNew + } +- s := make([]byte, priv.Size()) +- return c.FillBytes(s), nil ++ ++ return decryptAndCheck(priv, em) + } + + const ( +@@ -269,7 +275,7 @@ func SignPSS(rand io.Reader, priv *PrivateKey, hash crypto.Hash, digest []byte, + saltLength := opts.saltLength() + switch saltLength { + case PSSSaltLengthAuto: +- saltLength = (priv.N.BitLen()-1+7)/8 - 2 - hash.Size() ++ saltLength = (bigBitLen(priv.N)-1+7)/8 - 2 - hash.Size() + case PSSSaltLengthEqualsHash: + saltLength = hash.Size() + } +@@ -278,7 +284,7 @@ func SignPSS(rand io.Reader, priv *PrivateKey, hash crypto.Hash, digest []byte, + if _, err := io.ReadFull(rand, salt); err != nil { + return nil, err + } +- return signPSSWithSalt(rand, priv, hash, digest, salt) ++ return signPSSWithSalt(priv, hash, digest, salt) + } + + // VerifyPSS verifies a PSS signature. +@@ -291,13 +297,22 @@ func VerifyPSS(pub *PublicKey, hash crypto.Hash, digest []byte, sig []byte, opts + if len(sig) != pub.Size() { + return ErrVerification + } +- s := new(big.Int).SetBytes(sig) +- m := encrypt(new(big.Int), pub, s) +- emBits := pub.N.BitLen() - 1 ++ ++ emBits := bigBitLen(pub.N) - 1 + emLen := (emBits + 7) / 8 +- if m.BitLen() > emLen*8 { +- return ErrVerification ++ em := encrypt(pub, sig) ++ ++ // Like in signPSSWithSalt, deal with mismatches between emLen and the size ++ // of the modulus. The spec would have us wire emLen into the encoding ++ // function, but we'd rather always encode to the size of the modulus and ++ // then strip leading zeroes if necessary. This only happens for weird ++ // modulus sizes anyway. ++ for len(em) > emLen && len(em) > 0 { ++ if em[0] != 0 { ++ return ErrVerification ++ } ++ em = em[1:] + } +- em := m.FillBytes(make([]byte, emLen)) ++ + return emsaPSSVerify(digest, em, emBits, opts.saltLength(), hash.New()) + } +diff --git a/src/crypto/rsa/pss_test.go b/src/crypto/rsa/pss_test.go +index c3a6d46..d018b43 100644 +--- a/src/crypto/rsa/pss_test.go ++++ b/src/crypto/rsa/pss_test.go +@@ -233,7 +233,10 @@ func TestPSSSigning(t *testing.T) { + } + } + +-func TestSignWithPSSSaltLengthAuto(t *testing.T) { ++func TestPSS513(t *testing.T) { ++ // See Issue 42741, and separately, RFC 8017: "Note that the octet length of ++ // EM will be one less than k if modBits - 1 is divisible by 8 and equal to ++ // k otherwise, where k is the length in octets of the RSA modulus n." + key, err := GenerateKey(rand.Reader, 513) + if err != nil { + t.Fatal(err) +@@ -246,8 +249,9 @@ func TestSignWithPSSSaltLengthAuto(t *testing.T) { + if err != nil { + t.Fatal(err) + } +- if len(signature) == 0 { +- t.Fatal("empty signature returned") ++ err = VerifyPSS(&key.PublicKey, crypto.SHA256, digest[:], signature, nil) ++ if err != nil { ++ t.Error(err) + } + } + +diff --git a/src/crypto/rsa/rsa.go b/src/crypto/rsa/rsa.go +index 6fd59b3..20c1fe1 100644 +--- a/src/crypto/rsa/rsa.go ++++ b/src/crypto/rsa/rsa.go +@@ -19,13 +19,17 @@ + // over the public key primitive, the PrivateKey type implements the + // Decrypter and Signer interfaces from the crypto package. + // +-// The RSA operations in this package are not implemented using constant-time algorithms. ++// Operations in this package are implemented using constant-time algorithms, ++// except for [GenerateKey], [PrivateKey.Precompute], and [PrivateKey.Validate]. ++// Every other operation only leaks the bit size of the involved values, which ++// all depend on the selected key size. + package rsa + + import ( + "crypto" + "crypto/rand" + "crypto/subtle" ++ "encoding/binary" + "errors" + "hash" + "io" +@@ -35,7 +39,6 @@ import ( + "crypto/internal/randutil" + ) + +-var bigZero = big.NewInt(0) + var bigOne = big.NewInt(1) + + // A PublicKey represents the public part of an RSA key. +@@ -50,7 +53,7 @@ type PublicKey struct { + // Size returns the modulus size in bytes. Raw signatures and ciphertexts + // for or by this public key will have the same size. + func (pub *PublicKey) Size() int { +- return (pub.N.BitLen() + 7) / 8 ++ return (bigBitLen(pub.N) + 7) / 8 + } + + // Equal reports whether pub and x have the same value. +@@ -384,10 +387,18 @@ func mgf1XOR(out []byte, hash hash.Hash, seed []byte) { + // too large for the size of the public key. + var ErrMessageTooLong = errors.New("crypto/rsa: message too long for RSA public key size") + +-func encrypt(c *big.Int, pub *PublicKey, m *big.Int) *big.Int { +- e := big.NewInt(int64(pub.E)) +- c.Exp(m, e, pub.N) +- return c ++func encrypt(pub *PublicKey, plaintext []byte) []byte { ++ N := modulusFromNat(natFromBig(pub.N)) ++ m := natFromBytes(plaintext).expandFor(N) ++ ++ e := make([]byte, 8) ++ binary.BigEndian.PutUint64(e, uint64(pub.E)) ++ for len(e) > 1 && e[0] == 0 { ++ e = e[1:] ++ } ++ ++ out := make([]byte, modulusSize(N)) ++ return new(nat).exp(m, e, N).fillBytes(out) + } + + // EncryptOAEP encrypts the given message with RSA-OAEP. +@@ -437,12 +448,7 @@ func EncryptOAEP(hash hash.Hash, random io.Reader, pub *PublicKey, msg []byte, l + mgf1XOR(db, hash, seed) + mgf1XOR(seed, hash, db) + +- m := new(big.Int) +- m.SetBytes(em) +- c := encrypt(new(big.Int), pub, m) +- +- out := make([]byte, k) +- return c.FillBytes(out), nil ++ return encrypt(pub, em), nil + } + + // ErrDecryption represents a failure to decrypt a message. +@@ -484,98 +490,70 @@ func (priv *PrivateKey) Precompute() { + } + } + +-// decrypt performs an RSA decryption, resulting in a plaintext integer. If a +-// random source is given, RSA blinding is used. +-func decrypt(random io.Reader, priv *PrivateKey, c *big.Int) (m *big.Int, err error) { +- // TODO(agl): can we get away with reusing blinds? +- if c.Cmp(priv.N) > 0 { +- err = ErrDecryption +- return ++// decrypt performs an RSA decryption of ciphertext into out. ++func decrypt(priv *PrivateKey, ciphertext []byte) ([]byte, error) { ++ N := modulusFromNat(natFromBig(priv.N)) ++ c := natFromBytes(ciphertext).expandFor(N) ++ if c.cmpGeq(N.nat) == 1 { ++ return nil, ErrDecryption + } + if priv.N.Sign() == 0 { + return nil, ErrDecryption + } + +- var ir *big.Int +- if random != nil { +- randutil.MaybeReadByte(random) +- +- // Blinding enabled. Blinding involves multiplying c by r^e. +- // Then the decryption operation performs (m^e * r^e)^d mod n +- // which equals mr mod n. The factor of r can then be removed +- // by multiplying by the multiplicative inverse of r. +- +- var r *big.Int +- ir = new(big.Int) +- for { +- r, err = rand.Int(random, priv.N) +- if err != nil { +- return +- } +- if r.Cmp(bigZero) == 0 { +- r = bigOne +- } +- ok := ir.ModInverse(r, priv.N) +- if ok != nil { +- break +- } +- } +- bigE := big.NewInt(int64(priv.E)) +- rpowe := new(big.Int).Exp(r, bigE, priv.N) // N != 0 +- cCopy := new(big.Int).Set(c) +- cCopy.Mul(cCopy, rpowe) +- cCopy.Mod(cCopy, priv.N) +- c = cCopy +- } +- ++ // Note that because our private decryption exponents are stored as big.Int, ++ // we potentially leak the exact number of bits of these exponents. This ++ // isn't great, but should be fine. + if priv.Precomputed.Dp == nil { +- m = new(big.Int).Exp(c, priv.D, priv.N) +- } else { +- // We have the precalculated values needed for the CRT. +- m = new(big.Int).Exp(c, priv.Precomputed.Dp, priv.Primes[0]) +- m2 := new(big.Int).Exp(c, priv.Precomputed.Dq, priv.Primes[1]) +- m.Sub(m, m2) +- if m.Sign() < 0 { +- m.Add(m, priv.Primes[0]) +- } +- m.Mul(m, priv.Precomputed.Qinv) +- m.Mod(m, priv.Primes[0]) +- m.Mul(m, priv.Primes[1]) +- m.Add(m, m2) +- +- for i, values := range priv.Precomputed.CRTValues { +- prime := priv.Primes[2+i] +- m2.Exp(c, values.Exp, prime) +- m2.Sub(m2, m) +- m2.Mul(m2, values.Coeff) +- m2.Mod(m2, prime) +- if m2.Sign() < 0 { +- m2.Add(m2, prime) +- } +- m2.Mul(m2, values.R) +- m.Add(m, m2) +- } +- } +- +- if ir != nil { +- // Unblind. +- m.Mul(m, ir) +- m.Mod(m, priv.N) +- } +- +- return ++ out := make([]byte, modulusSize(N)) ++ return new(nat).exp(c, priv.D.Bytes(), N).fillBytes(out), nil ++ } ++ ++ t0 := new(nat) ++ P := modulusFromNat(natFromBig(priv.Primes[0])) ++ Q := modulusFromNat(natFromBig(priv.Primes[1])) ++ // m = c ^ Dp mod p ++ m := new(nat).exp(t0.mod(c, P), priv.Precomputed.Dp.Bytes(), P) ++ // m2 = c ^ Dq mod q ++ m2 := new(nat).exp(t0.mod(c, Q), priv.Precomputed.Dq.Bytes(), Q) ++ // m = m - m2 mod p ++ m.modSub(t0.mod(m2, P), P) ++ // m = m * Qinv mod p ++ m.modMul(natFromBig(priv.Precomputed.Qinv).expandFor(P), P) ++ // m = m * q mod N ++ m.expandFor(N).modMul(t0.mod(Q.nat, N), N) ++ // m = m + m2 mod N ++ m.modAdd(m2.expandFor(N), N) ++ ++ for i, values := range priv.Precomputed.CRTValues { ++ p := modulusFromNat(natFromBig(priv.Primes[2+i])) ++ // m2 = c ^ Exp mod p ++ m2.exp(t0.mod(c, p), values.Exp.Bytes(), p) ++ // m2 = m2 - m mod p ++ m2.modSub(t0.mod(m, p), p) ++ // m2 = m2 * Coeff mod p ++ m2.modMul(natFromBig(values.Coeff).expandFor(p), p) ++ // m2 = m2 * R mod N ++ R := natFromBig(values.R).expandFor(N) ++ m2.expandFor(N).modMul(R, N) ++ // m = m + m2 mod N ++ m.modAdd(m2, N) ++ } ++ ++ out := make([]byte, modulusSize(N)) ++ return m.fillBytes(out), nil + } + +-func decryptAndCheck(random io.Reader, priv *PrivateKey, c *big.Int) (m *big.Int, err error) { +- m, err = decrypt(random, priv, c) ++func decryptAndCheck(priv *PrivateKey, ciphertext []byte) (m []byte, err error) { ++ m, err = decrypt(priv, ciphertext) + if err != nil { + return nil, err + } + + // In order to defend against errors in the CRT computation, m^e is + // calculated, which should match the original ciphertext. +- check := encrypt(new(big.Int), &priv.PublicKey, m) +- if c.Cmp(check) != 0 { ++ check := encrypt(&priv.PublicKey, m) ++ if subtle.ConstantTimeCompare(ciphertext, check) != 1 { + return nil, errors.New("rsa: internal error") + } + return m, nil +@@ -587,9 +565,7 @@ func decryptAndCheck(random io.Reader, priv *PrivateKey, c *big.Int) (m *big.Int + // Encryption and decryption of a given message must use the same hash function + // and sha256.New() is a reasonable choice. + // +-// The random parameter, if not nil, is used to blind the private-key operation +-// and avoid timing side-channel attacks. Blinding is purely internal to this +-// function – the random data need not match that used when encrypting. ++// The random parameter is legacy and ignored, and it can be as nil. + // + // The label parameter must match the value given when encrypting. See + // EncryptOAEP for details. +@@ -603,9 +579,7 @@ func DecryptOAEP(hash hash.Hash, random io.Reader, priv *PrivateKey, ciphertext + return nil, ErrDecryption + } + +- c := new(big.Int).SetBytes(ciphertext) +- +- m, err := decrypt(random, priv, c) ++ em, err := decrypt(priv, ciphertext) + if err != nil { + return nil, err + } +@@ -614,10 +588,6 @@ func DecryptOAEP(hash hash.Hash, random io.Reader, priv *PrivateKey, ciphertext + lHash := hash.Sum(nil) + hash.Reset() + +- // We probably leak the number of leading zeros. +- // It's not clear that we can do anything about this. +- em := m.FillBytes(make([]byte, k)) +- + firstByteIsZero := subtle.ConstantTimeByteEq(em[0], 0) + + seed := em[1 : hash.Size()+1] +-- +2.40.0 -- 2.34.1 ^ permalink raw reply related [flat|nested] 33+ messages in thread
* [OE-core][kirkstone 3/8] curl: Fix CVE-2023-46219 2024-02-12 13:54 [OE-core][kirkstone 0/8] Patch review Steve Sakoman 2024-02-12 13:54 ` [OE-core][kirkstone 1/8] libxml2: Fix for CVE-2024-25062 Steve Sakoman 2024-02-12 13:54 ` [OE-core][kirkstone 2/8] go: Fix CVE-2023-45285 and CVE-2023-45287 Steve Sakoman @ 2024-02-12 13:54 ` Steve Sakoman 2024-02-12 13:54 ` [OE-core][kirkstone 4/8] libgit2: Fix CVE-2024-24575 and CVE-2024-24577 Steve Sakoman ` (4 subsequent siblings) 7 siblings, 0 replies; 33+ messages in thread From: Steve Sakoman @ 2024-02-12 13:54 UTC (permalink / raw) To: openembedded-core From: Archana Polampalli <archana.polampalli@windriver.com> When saving HSTS data to an excessively long file name, curl could end up removing all contents, making subsequent requests using that file unaware of the HSTS status they should otherwise use. Signed-off-by: Archana Polampalli <archana.polampalli@windriver.com> Signed-off-by: Steve Sakoman <steve@sakoman.com> --- .../curl/curl/CVE-2023-46219-0001.patch | 42 ++++++ .../curl/curl/CVE-2023-46219-0002.patch | 133 ++++++++++++++++++ .../curl/curl/CVE-2023-46219-0003.patch | 81 +++++++++++ meta/recipes-support/curl/curl_7.82.0.bb | 3 + 4 files changed, 259 insertions(+) create mode 100644 meta/recipes-support/curl/curl/CVE-2023-46219-0001.patch create mode 100644 meta/recipes-support/curl/curl/CVE-2023-46219-0002.patch create mode 100644 meta/recipes-support/curl/curl/CVE-2023-46219-0003.patch diff --git a/meta/recipes-support/curl/curl/CVE-2023-46219-0001.patch b/meta/recipes-support/curl/curl/CVE-2023-46219-0001.patch new file mode 100644 index 0000000000..55e8f6fac9 --- /dev/null +++ b/meta/recipes-support/curl/curl/CVE-2023-46219-0001.patch @@ -0,0 +1,42 @@ +From 0c667188e0c6cda615a036b8a2b4125f2c404dde Mon Sep 17 00:00:00 2001 +From: SaltyMilk <soufiane.elmelcaoui@gmail.com> +Date: Mon, 10 Jul 2023 21:43:28 +0200 +Subject: [PATCH] fopen: optimize + +Closes #11419 + +CVE: CVE-2023-46219 + +Upstream-Status: Backport [https://github.com/curl/curl/commit/0c667188e0c6] + +Signed-off-by: Archana Polampalli <archana.polampalli@windriver.com> +--- + lib/fopen.c | 12 ++++++------ + 1 file changed, 6 insertions(+), 6 deletions(-) + +diff --git a/lib/fopen.c b/lib/fopen.c +index ad3691b..92f39cf 100644 +--- a/lib/fopen.c ++++ b/lib/fopen.c +@@ -56,13 +56,13 @@ CURLcode Curl_fopen(struct Curl_easy *data, const char *filename, + int fd = -1; + *tempname = NULL; + +- if(stat(filename, &sb) == -1 || !S_ISREG(sb.st_mode)) { +- /* a non-regular file, fallback to direct fopen() */ +- *fh = fopen(filename, FOPEN_WRITETEXT); +- if(*fh) +- return CURLE_OK; ++ *fh = fopen(filename, FOPEN_WRITETEXT); ++ if(!*fh) + goto fail; +- } ++ if(fstat(fileno(*fh), &sb) == -1 || !S_ISREG(sb.st_mode)) ++ return CURLE_OK; ++ fclose(*fh); ++ *fh = NULL; + + result = Curl_rand_hex(data, randsuffix, sizeof(randsuffix)); + if(result) +-- +2.40.0 diff --git a/meta/recipes-support/curl/curl/CVE-2023-46219-0002.patch b/meta/recipes-support/curl/curl/CVE-2023-46219-0002.patch new file mode 100644 index 0000000000..f432fabbb1 --- /dev/null +++ b/meta/recipes-support/curl/curl/CVE-2023-46219-0002.patch @@ -0,0 +1,133 @@ +From 73b65e94f3531179de45c6f3c836a610e3d0a846 Mon Sep 17 00:00:00 2001 +From: Daniel Stenberg <daniel@haxx.se> +Date: Thu, 23 Nov 2023 08:23:17 +0100 +Subject: [PATCH] fopen: create short(er) temporary file name + +Only using random letters in the name plus a ".tmp" extension. Not by +appending characters to the final file name. + +Reported-by: Maksymilian Arciemowicz + +Closes #12388 + +CVE: CVE-2023-46219 + +Upstream-Status: Backport [https://github.com/curl/curl/commit/73b65e94f3531179] + +Signed-off-by: Archana Polampalli <archana.polampalli@windriver.com> +--- + lib/fopen.c | 63 +++++++++++++++++++++++++++++++++++++++++++++++++---- + 1 file changed, 59 insertions(+), 4 deletions(-) + +diff --git a/lib/fopen.c b/lib/fopen.c +index 92f39cf..1670e32 100644 +--- a/lib/fopen.c ++++ b/lib/fopen.c +@@ -39,6 +39,50 @@ + #include "curl_memory.h" + #include "memdebug.h" + ++ ++/* ++ The dirslash() function breaks a null-terminated pathname string into ++ directory and filename components then returns the directory component up ++ to, *AND INCLUDING*, a final '/'. If there is no directory in the path, ++ this instead returns a "" string. ++ This function returns a pointer to malloc'ed memory. ++ The input path to this function is expected to have a file name part. ++*/ ++ ++#ifdef _WIN32 ++#define PATHSEP "\\" ++#define IS_SEP(x) (((x) == '/') || ((x) == '\\')) ++#elif defined(MSDOS) || defined(__EMX__) || defined(OS2) ++#define PATHSEP "\\" ++#define IS_SEP(x) ((x) == '\\') ++#else ++#define PATHSEP "/" ++#define IS_SEP(x) ((x) == '/') ++#endif ++ ++static char *dirslash(const char *path) ++{ ++ size_t n; ++ struct dynbuf out; ++ DEBUGASSERT(path); ++ Curl_dyn_init(&out, CURL_MAX_INPUT_LENGTH); ++ n = strlen(path); ++ if(n) { ++ /* find the rightmost path separator, if any */ ++ while(n && !IS_SEP(path[n-1])) ++ --n; ++ /* skip over all the path separators, if any */ ++ while(n && IS_SEP(path[n-1])) ++ --n; ++ } ++ if(Curl_dyn_addn(&out, path, n)) ++ return NULL; ++ /* if there was a directory, append a single trailing slash */ ++ if(n && Curl_dyn_addn(&out, PATHSEP, 1)) ++ return NULL; ++ return Curl_dyn_ptr(&out); ++} ++ + /* + * Curl_fopen() opens a file for writing with a temp name, to be renamed + * to the final name when completed. If there is an existing file using this +@@ -50,25 +94,34 @@ CURLcode Curl_fopen(struct Curl_easy *data, const char *filename, + FILE **fh, char **tempname) + { + CURLcode result = CURLE_WRITE_ERROR; +- unsigned char randsuffix[9]; ++ unsigned char randbuf[41]; + char *tempstore = NULL; + struct_stat sb; + int fd = -1; ++ char *dir; + *tempname = NULL; + ++ dir = dirslash(filename); ++ if(!dir) ++ goto fail; ++ + *fh = fopen(filename, FOPEN_WRITETEXT); + if(!*fh) + goto fail; +- if(fstat(fileno(*fh), &sb) == -1 || !S_ISREG(sb.st_mode)) ++ if(fstat(fileno(*fh), &sb) == -1 || !S_ISREG(sb.st_mode)){ ++ free(dir); + return CURLE_OK; ++ } + fclose(*fh); + *fh = NULL; + +- result = Curl_rand_hex(data, randsuffix, sizeof(randsuffix)); ++ result = Curl_rand_hex(data, randbuf, sizeof(randbuf)); + if(result) + goto fail; + +- tempstore = aprintf("%s.%s.tmp", filename, randsuffix); ++ /* The temp file name should not end up too long for the target file ++ system */ ++ tempstore = aprintf("%s%s.tmp", dir, randbuf); + if(!tempstore) { + result = CURLE_OUT_OF_MEMORY; + goto fail; +@@ -95,6 +148,7 @@ CURLcode Curl_fopen(struct Curl_easy *data, const char *filename, + if(!*fh) + goto fail; + ++ free(dir); + *tempname = tempstore; + return CURLE_OK; + +@@ -107,6 +161,7 @@ fail: + free(tempstore); + + *tempname = NULL; ++ free(dir); + return result; + } + +-- +2.40.0 diff --git a/meta/recipes-support/curl/curl/CVE-2023-46219-0003.patch b/meta/recipes-support/curl/curl/CVE-2023-46219-0003.patch new file mode 100644 index 0000000000..3b6f756549 --- /dev/null +++ b/meta/recipes-support/curl/curl/CVE-2023-46219-0003.patch @@ -0,0 +1,81 @@ +From f27b8dba73295cb5296a50f2c19c0739b502eb94 Mon Sep 17 00:00:00 2001 +From: Daniel Stenberg <daniel@haxx.se> +Date: Fri, 24 Nov 2023 09:46:32 +0100 +Subject: [PATCH] fopen: allocate the dir after fopen + +Move the allocation of the directory name down to after the fopen() call +to allow that shortcut code path to avoid a superfluous malloc+free +cycle. + +Follow-up to 73b65e94f35311 + +Closes #12398 + +CVE: CVE-2023-46219 + +Upstream-Status: Backport [https://github.com/curl/curl/commit/f27b8dba73295cb529] + +Signed-off-by: Archana Polampalli <archana.polampalli@windriver.com> +--- + lib/fopen.c | 19 ++++++++----------- + 1 file changed, 8 insertions(+), 11 deletions(-) + +diff --git a/lib/fopen.c b/lib/fopen.c +index 1670e32..b663f8b 100644 +--- a/lib/fopen.c ++++ b/lib/fopen.c +@@ -98,18 +98,13 @@ CURLcode Curl_fopen(struct Curl_easy *data, const char *filename, + char *tempstore = NULL; + struct_stat sb; + int fd = -1; +- char *dir; ++ char *dir = NULL; + *tempname = NULL; + +- dir = dirslash(filename); +- if(!dir) +- goto fail; +- + *fh = fopen(filename, FOPEN_WRITETEXT); + if(!*fh) + goto fail; + if(fstat(fileno(*fh), &sb) == -1 || !S_ISREG(sb.st_mode)){ +- free(dir); + return CURLE_OK; + } + fclose(*fh); +@@ -119,9 +114,13 @@ CURLcode Curl_fopen(struct Curl_easy *data, const char *filename, + if(result) + goto fail; + +- /* The temp file name should not end up too long for the target file +- system */ +- tempstore = aprintf("%s%s.tmp", dir, randbuf); ++ dir = dirslash(filename); ++ if(dir) { ++ /* The temp file name should not end up too long for the target file ++ system */ ++ tempstore = aprintf("%s%s.tmp", dir, randbuf); ++ free(dir); ++ } + if(!tempstore) { + result = CURLE_OUT_OF_MEMORY; + goto fail; +@@ -148,7 +147,6 @@ CURLcode Curl_fopen(struct Curl_easy *data, const char *filename, + if(!*fh) + goto fail; + +- free(dir); + *tempname = tempstore; + return CURLE_OK; + +@@ -161,7 +159,6 @@ fail: + free(tempstore); + + *tempname = NULL; +- free(dir); + return result; + } + +-- +2.40.0 diff --git a/meta/recipes-support/curl/curl_7.82.0.bb b/meta/recipes-support/curl/curl_7.82.0.bb index 965f05bc98..de69d3d53b 100644 --- a/meta/recipes-support/curl/curl_7.82.0.bb +++ b/meta/recipes-support/curl/curl_7.82.0.bb @@ -54,6 +54,9 @@ SRC_URI = "https://curl.se/download/${BP}.tar.xz \ file://CVE-2023-38545.patch \ file://CVE-2023-38546.patch \ file://CVE-2023-46218.patch \ + file://CVE-2023-46219-0001.patch \ + file://CVE-2023-46219-0002.patch \ + file://CVE-2023-46219-0003.patch \ " SRC_URI[sha256sum] = "0aaa12d7bd04b0966254f2703ce80dd5c38dbbd76af0297d3d690cdce58a583c" -- 2.34.1 ^ permalink raw reply related [flat|nested] 33+ messages in thread
* [OE-core][kirkstone 4/8] libgit2: Fix CVE-2024-24575 and CVE-2024-24577 2024-02-12 13:54 [OE-core][kirkstone 0/8] Patch review Steve Sakoman ` (2 preceding siblings ...) 2024-02-12 13:54 ` [OE-core][kirkstone 3/8] curl: Fix CVE-2023-46219 Steve Sakoman @ 2024-02-12 13:54 ` Steve Sakoman 2024-02-12 13:54 ` [OE-core][kirkstone 5/8] python3-pycryptodome: Fix CVE-2023-52323 Steve Sakoman ` (3 subsequent siblings) 7 siblings, 0 replies; 33+ messages in thread From: Steve Sakoman @ 2024-02-12 13:54 UTC (permalink / raw) To: openembedded-core From: Soumya Sambu <soumya.sambu@windriver.com> CVE-2024-24575: libgit2 is a portable C implementation of the Git core methods provided as a linkable library with a solid API, allowing to build Git functionality into your application. Using well-crafted inputs to `git_revparse_single` can cause the function to enter an infinite loop, potentially causing a Denial of Service attack in the calling application. The revparse function in `src/libgit2/revparse.c` uses a loop to parse the user-provided spec string. There is an edge-case during parsing that allows a bad actor to force the loop conditions to access arbitrary memory. Potentially, this could also leak memory if the extracted rev spec is reflected back to the attacker. As such, libgit2 versions before 1.4.0 are not affected. Users should upgrade to version 1.6.5 or 1.7.2. CVE-2024-24577: libgit2 is a portable C implementation of the Git core methods provided as a linkable library with a solid API, allowing to build Git functionality into your application. Using well-crafted inputs to `git_index_add` can cause heap corruption that could be leveraged for arbitrary code execution. There is an issue in the `has_dir_name` function in `src/libgit2/index.c`, which frees an entry that should not be freed. The freed entry is later used and overwritten with potentially bad actor-controlled data leading to controlled heap corruption. Depending on the application that uses libgit2, this could lead to arbitrary code execution. This issue has been patched in version 1.6.5 and 1.7.2. References: https://nvd.nist.gov/vuln/detail/CVE-2024-24575 https://security-tracker.debian.org/tracker/CVE-2024-24575 https://nvd.nist.gov/vuln/detail/CVE-2024-24577 https://security-tracker.debian.org/tracker/CVE-2024-24577 Signed-off-by: Soumya Sambu <soumya.sambu@windriver.com> Signed-off-by: Steve Sakoman <steve@sakoman.com> --- .../libgit2/libgit2/CVE-2024-24575.patch | 56 +++++++++++++++++++ .../libgit2/libgit2/CVE-2024-24577.patch | 52 +++++++++++++++++ meta/recipes-support/libgit2/libgit2_1.4.5.bb | 5 +- 3 files changed, 112 insertions(+), 1 deletion(-) create mode 100644 meta/recipes-support/libgit2/libgit2/CVE-2024-24575.patch create mode 100644 meta/recipes-support/libgit2/libgit2/CVE-2024-24577.patch diff --git a/meta/recipes-support/libgit2/libgit2/CVE-2024-24575.patch b/meta/recipes-support/libgit2/libgit2/CVE-2024-24575.patch new file mode 100644 index 0000000000..d3957ac5d0 --- /dev/null +++ b/meta/recipes-support/libgit2/libgit2/CVE-2024-24575.patch @@ -0,0 +1,56 @@ +From c9d31b711e8906cf248566f43142f20b03e20cbf Mon Sep 17 00:00:00 2001 +From: Edward Thomson <ethomson@edwardthomson.com> +Date: Fri, 17 Nov 2023 16:54:47 +0000 +Subject: [PATCH] revparse: fix parsing bug for trailing `@` + +When parsing a revspec that ends with a trailing `@`, explicitly stop +parsing. Introduce a sentinel variable to explicitly stop parsing. + +Prior to this, we would set `spec` to `HEAD`, but were looping on the +value of `spec[pos]`, so we would continue walking the (new) `spec` +at offset `pos`, looking for a NUL. This is obviously an out-of-bounds +read. + +Credit to Michael Rodler (@f0rki) and Amazon AWS Security. + +CVE: CVE-2024-24575 + +Upstream-Status: Backport [https://github.com/libgit2/libgit2/commit/c9d31b711e8906cf248566f43142f20b03e20cbf] + +Signed-off-by: Soumya Sambu <soumya.sambu@windriver.com> +--- + src/revparse.c | 5 ++++- + 1 file changed, 4 insertions(+), 1 deletion(-) + +diff --git a/src/revparse.c b/src/revparse.c +index 9bc28e9fc..d3bbe840b 100644 +--- a/src/revparse.c ++++ b/src/revparse.c +@@ -685,6 +685,7 @@ static int revparse( + git_object *base_rev = NULL; + + bool should_return_reference = true; ++ bool parsed = false; + + GIT_ASSERT_ARG(object_out); + GIT_ASSERT_ARG(reference_out); +@@ -694,7 +695,7 @@ static int revparse( + *object_out = NULL; + *reference_out = NULL; + +- while (spec[pos]) { ++ while (!parsed && spec[pos]) { + switch (spec[pos]) { + case '^': + should_return_reference = false; +@@ -801,6 +802,8 @@ static int revparse( + break; + } else if (spec[pos+1] == '\0') { + spec = "HEAD"; ++ identifier_len = 4; ++ parsed = true; + break; + } + /* fall through */ +-- +2.40.0 diff --git a/meta/recipes-support/libgit2/libgit2/CVE-2024-24577.patch b/meta/recipes-support/libgit2/libgit2/CVE-2024-24577.patch new file mode 100644 index 0000000000..3469f9d099 --- /dev/null +++ b/meta/recipes-support/libgit2/libgit2/CVE-2024-24577.patch @@ -0,0 +1,52 @@ +From eb4c1716cd92bf56f2770653a915d5fc01eab8f3 Mon Sep 17 00:00:00 2001 +From: Edward Thomson <ethomson@edwardthomson.com> +Date: Sat, 16 Dec 2023 11:19:07 +0000 +Subject: [PATCH] index: correct index has_dir_name check + +`has_dir_name` is used to check for directory/file collisions, +and attempts to determine whether the index contains a file with +a directory name that is a proper subset of the new index entry +that we're trying to add. + +To determine directory name, the function would walk the path string +backwards to identify a `/`, stopping at the end of the string. However, +the function assumed that the strings did not start with a `/`. If the +paths contain only a single `/` at the beginning of the string, then the +function would continue the loop, erroneously, when they should have +stopped at the first character. + +Correct the order of the tests to terminate properly. + +Credit to Michael Rodler (@f0rki) and Amazon AWS Security. + +CVE: CVE-2024-24577 + +Upstream-Status: Backport [https://github.com/libgit2/libgit2/commit/eb4c1716cd92bf56f2770653a915d5fc01eab8f3] + +Signed-off-by: Soumya Sambu <soumya.sambu@windriver.com> +--- + src/index.c | 7 +++++-- + 1 file changed, 5 insertions(+), 2 deletions(-) + +diff --git a/src/index.c b/src/index.c +index aa97c6421..e8ff82e1a 100644 +--- a/src/index.c ++++ b/src/index.c +@@ -1148,10 +1148,13 @@ static int has_dir_name(git_index *index, + size_t len, pos; + + for (;;) { +- if (*--slash == '/') +- break; ++ slash--; ++ + if (slash <= entry->path) + return 0; ++ ++ if (*slash == '/') ++ break; + } + len = slash - name; + +-- +2.40.0 diff --git a/meta/recipes-support/libgit2/libgit2_1.4.5.bb b/meta/recipes-support/libgit2/libgit2_1.4.5.bb index aadfe4ad02..ad8b9a536a 100644 --- a/meta/recipes-support/libgit2/libgit2_1.4.5.bb +++ b/meta/recipes-support/libgit2/libgit2_1.4.5.bb @@ -5,7 +5,10 @@ LIC_FILES_CHKSUM = "file://COPYING;md5=e5a9227de4cb6afb5d35ed7b0fdf480d" DEPENDS = "curl openssl zlib libssh2 libgcrypt libpcre2" -SRC_URI = "git://github.com/libgit2/libgit2.git;branch=maint/v1.4;protocol=https" +SRC_URI = "git://github.com/libgit2/libgit2.git;branch=maint/v1.4;protocol=https \ + file://CVE-2024-24575.patch \ + file://CVE-2024-24577.patch \ + " SRCREV = "cd6f679af401eda1f172402006ef8265f8bd58ea" S = "${WORKDIR}/git" -- 2.34.1 ^ permalink raw reply related [flat|nested] 33+ messages in thread
* [OE-core][kirkstone 5/8] python3-pycryptodome: Fix CVE-2023-52323 2024-02-12 13:54 [OE-core][kirkstone 0/8] Patch review Steve Sakoman ` (3 preceding siblings ...) 2024-02-12 13:54 ` [OE-core][kirkstone 4/8] libgit2: Fix CVE-2024-24575 and CVE-2024-24577 Steve Sakoman @ 2024-02-12 13:54 ` Steve Sakoman 2024-02-12 13:54 ` [OE-core][kirkstone 6/8] systemd: Only add myhostname to nsswitch.conf if in PACKAGECONFIG Steve Sakoman ` (2 subsequent siblings) 7 siblings, 0 replies; 33+ messages in thread From: Steve Sakoman @ 2024-02-12 13:54 UTC (permalink / raw) To: openembedded-core From: Narpat Mali <narpat.mali@windriver.com> PyCryptodome and pycryptodomex before 3.19.1 allow side-channel leakage for OAEP decryption, exploitable for a Manger attack. References: https://security-tracker.debian.org/tracker/CVE-2023-52323 https://github.com/Legrandin/pycryptodome/blob/master/Changelog.rst Signed-off-by: Narpat Mali <narpat.mali@windriver.com> Signed-off-by: Steve Sakoman <steve@sakoman.com> --- .../python3-pycryptodome/CVE-2023-52323.patch | 436 ++++++++++++++++++ .../python/python3-pycryptodome_3.14.1.bb | 1 + .../CVE-2023-52323.patch | 436 ++++++++++++++++++ .../python/python3-pycryptodomex_3.14.1.bb | 2 + 4 files changed, 875 insertions(+) create mode 100644 meta/recipes-devtools/python/python3-pycryptodome/CVE-2023-52323.patch create mode 100644 meta/recipes-devtools/python/python3-pycryptodomex/CVE-2023-52323.patch diff --git a/meta/recipes-devtools/python/python3-pycryptodome/CVE-2023-52323.patch b/meta/recipes-devtools/python/python3-pycryptodome/CVE-2023-52323.patch new file mode 100644 index 0000000000..be3090eb8d --- /dev/null +++ b/meta/recipes-devtools/python/python3-pycryptodome/CVE-2023-52323.patch @@ -0,0 +1,436 @@ +From 73bbed822fadddf3c0ab4a945ee6ab16bbca6961 Mon Sep 17 00:00:00 2001 +From: Helder Eijs <helderijs@gmail.com> +Date: Thu, 1 Feb 2024 13:43:44 +0000 +Subject: [PATCH] Use constant-time (faster) padding decoding also for OAEP + +CVE: CVE-2023-52323 + +Upstream-Status: Backport [https://github.com/Legrandin/pycryptodome/commit/0deea1bfe1489e8c80d2053bbb06a1aa0b181ebd] + +Signed-off-by: Narpat Mali <narpat.mali@windriver.com> +--- + lib/Crypto/Cipher/PKCS1_OAEP.py | 38 +++++------- + lib/Crypto/Cipher/PKCS1_v1_5.py | 31 +--------- + lib/Crypto/Cipher/_pkcs1_oaep_decode.py | 41 +++++++++++++ + src/pkcs1_decode.c | 79 +++++++++++++++++++++++-- + src/test/test_pkcs1.c | 22 +++---- + 5 files changed, 145 insertions(+), 66 deletions(-) + create mode 100644 lib/Crypto/Cipher/_pkcs1_oaep_decode.py + +diff --git a/lib/Crypto/Cipher/PKCS1_OAEP.py b/lib/Crypto/Cipher/PKCS1_OAEP.py +index 57a982b..6974584 100644 +--- a/lib/Crypto/Cipher/PKCS1_OAEP.py ++++ b/lib/Crypto/Cipher/PKCS1_OAEP.py +@@ -23,11 +23,13 @@ + from Crypto.Signature.pss import MGF1 + import Crypto.Hash.SHA1 + +-from Crypto.Util.py3compat import bord, _copy_bytes ++from Crypto.Util.py3compat import _copy_bytes + import Crypto.Util.number +-from Crypto.Util.number import ceil_div, bytes_to_long, long_to_bytes +-from Crypto.Util.strxor import strxor ++from Crypto.Util.number import ceil_div, bytes_to_long, long_to_bytes ++from Crypto.Util.strxor import strxor + from Crypto import Random ++from ._pkcs1_oaep_decode import oaep_decode ++ + + class PKCS1OAEP_Cipher: + """Cipher object for PKCS#1 v1.5 OAEP. +@@ -68,7 +70,7 @@ class PKCS1OAEP_Cipher: + if mgfunc: + self._mgf = mgfunc + else: +- self._mgf = lambda x,y: MGF1(x,y,self._hashObj) ++ self._mgf = lambda x, y: MGF1(x, y, self._hashObj) + + self._label = _copy_bytes(None, None, label) + self._randfunc = randfunc +@@ -105,7 +107,7 @@ class PKCS1OAEP_Cipher: + + # See 7.1.1 in RFC3447 + modBits = Crypto.Util.number.size(self._key.n) +- k = ceil_div(modBits, 8) # Convert from bits to bytes ++ k = ceil_div(modBits, 8) # Convert from bits to bytes + hLen = self._hashObj.digest_size + mLen = len(message) + +@@ -159,11 +161,11 @@ class PKCS1OAEP_Cipher: + + # See 7.1.2 in RFC3447 + modBits = Crypto.Util.number.size(self._key.n) +- k = ceil_div(modBits,8) # Convert from bits to bytes ++ k = ceil_div(modBits, 8) # Convert from bits to bytes + hLen = self._hashObj.digest_size + + # Step 1b and 1c +- if len(ciphertext) != k or k<hLen+2: ++ if len(ciphertext) != k or k < hLen+2: + raise ValueError("Ciphertext with incorrect length.") + # Step 2a (O2SIP) + ct_int = bytes_to_long(ciphertext) +@@ -173,8 +175,6 @@ class PKCS1OAEP_Cipher: + em = long_to_bytes(m_int, k) + # Step 3a + lHash = self._hashObj.new(self._label).digest() +- # Step 3b +- y = em[0] + # y must be 0, but we MUST NOT check it here in order not to + # allow attacks like Manger's (http://dl.acm.org/citation.cfm?id=704143) + maskedSeed = em[1:hLen+1] +@@ -187,22 +187,17 @@ class PKCS1OAEP_Cipher: + dbMask = self._mgf(seed, k-hLen-1) + # Step 3f + db = strxor(maskedDB, dbMask) +- # Step 3g +- one_pos = hLen + db[hLen:].find(b'\x01') +- lHash1 = db[:hLen] +- invalid = bord(y) | int(one_pos < hLen) +- hash_compare = strxor(lHash1, lHash) +- for x in hash_compare: +- invalid |= bord(x) +- for x in db[hLen:one_pos]: +- invalid |= bord(x) +- if invalid != 0: ++ # Step 3b + 3g ++ res = oaep_decode(em, lHash, db) ++ if res <= 0: + raise ValueError("Incorrect decryption.") + # Step 4 +- return db[one_pos + 1:] ++ return db[res:] ++ + + def new(key, hashAlgo=None, mgfunc=None, label=b'', randfunc=None): +- """Return a cipher object :class:`PKCS1OAEP_Cipher` that can be used to perform PKCS#1 OAEP encryption or decryption. ++ """Return a cipher object :class:`PKCS1OAEP_Cipher` ++ that can be used to perform PKCS#1 OAEP encryption or decryption. + + :param key: + The key object to use to encrypt or decrypt the message. +@@ -236,4 +231,3 @@ def new(key, hashAlgo=None, mgfunc=None, label=b'', randfunc=None): + if randfunc is None: + randfunc = Random.get_random_bytes + return PKCS1OAEP_Cipher(key, hashAlgo, mgfunc, label, randfunc) +- +diff --git a/lib/Crypto/Cipher/PKCS1_v1_5.py b/lib/Crypto/Cipher/PKCS1_v1_5.py +index d0d474a..94e99cf 100644 +--- a/lib/Crypto/Cipher/PKCS1_v1_5.py ++++ b/lib/Crypto/Cipher/PKCS1_v1_5.py +@@ -25,31 +25,7 @@ __all__ = ['new', 'PKCS115_Cipher'] + from Crypto import Random + from Crypto.Util.number import bytes_to_long, long_to_bytes + from Crypto.Util.py3compat import bord, is_bytes, _copy_bytes +- +-from Crypto.Util._raw_api import (load_pycryptodome_raw_lib, c_size_t, +- c_uint8_ptr) +- +- +-_raw_pkcs1_decode = load_pycryptodome_raw_lib("Crypto.Cipher._pkcs1_decode", +- """ +- int pkcs1_decode(const uint8_t *em, size_t len_em, +- const uint8_t *sentinel, size_t len_sentinel, +- size_t expected_pt_len, +- uint8_t *output); +- """) +- +- +-def _pkcs1_decode(em, sentinel, expected_pt_len, output): +- if len(em) != len(output): +- raise ValueError("Incorrect output length") +- +- ret = _raw_pkcs1_decode.pkcs1_decode(c_uint8_ptr(em), +- c_size_t(len(em)), +- c_uint8_ptr(sentinel), +- c_size_t(len(sentinel)), +- c_size_t(expected_pt_len), +- c_uint8_ptr(output)) +- return ret ++from ._pkcs1_oaep_decode import pkcs1_decode + + + class PKCS115_Cipher: +@@ -113,7 +89,6 @@ class PKCS115_Cipher: + continue + ps.append(new_byte) + ps = b"".join(ps) +- assert(len(ps) == k - mLen - 3) + # Step 2b + em = b'\x00\x02' + ps + b'\x00' + _copy_bytes(None, None, message) + # Step 3a (OS2IP) +@@ -185,14 +160,14 @@ class PKCS115_Cipher: + # Step 3 (not constant time when the sentinel is not a byte string) + output = bytes(bytearray(k)) + if not is_bytes(sentinel) or len(sentinel) > k: +- size = _pkcs1_decode(em, b'', expected_pt_len, output) ++ size = pkcs1_decode(em, b'', expected_pt_len, output) + if size < 0: + return sentinel + else: + return output[size:] + + # Step 3 (somewhat constant time) +- size = _pkcs1_decode(em, sentinel, expected_pt_len, output) ++ size = pkcs1_decode(em, sentinel, expected_pt_len, output) + return output[size:] + + +diff --git a/lib/Crypto/Cipher/_pkcs1_oaep_decode.py b/lib/Crypto/Cipher/_pkcs1_oaep_decode.py +new file mode 100644 +index 0000000..fc07528 +--- /dev/null ++++ b/lib/Crypto/Cipher/_pkcs1_oaep_decode.py +@@ -0,0 +1,41 @@ ++from Crypto.Util._raw_api import (load_pycryptodome_raw_lib, c_size_t, ++ c_uint8_ptr) ++ ++ ++_raw_pkcs1_decode = load_pycryptodome_raw_lib("Crypto.Cipher._pkcs1_decode", ++ """ ++ int pkcs1_decode(const uint8_t *em, size_t len_em, ++ const uint8_t *sentinel, size_t len_sentinel, ++ size_t expected_pt_len, ++ uint8_t *output); ++ ++ int oaep_decode(const uint8_t *em, ++ size_t em_len, ++ const uint8_t *lHash, ++ size_t hLen, ++ const uint8_t *db, ++ size_t db_len); ++ """) ++ ++ ++def pkcs1_decode(em, sentinel, expected_pt_len, output): ++ if len(em) != len(output): ++ raise ValueError("Incorrect output length") ++ ++ ret = _raw_pkcs1_decode.pkcs1_decode(c_uint8_ptr(em), ++ c_size_t(len(em)), ++ c_uint8_ptr(sentinel), ++ c_size_t(len(sentinel)), ++ c_size_t(expected_pt_len), ++ c_uint8_ptr(output)) ++ return ret ++ ++ ++def oaep_decode(em, lHash, db): ++ ret = _raw_pkcs1_decode.oaep_decode(c_uint8_ptr(em), ++ c_size_t(len(em)), ++ c_uint8_ptr(lHash), ++ c_size_t(len(lHash)), ++ c_uint8_ptr(db), ++ c_size_t(len(db))) ++ return ret +diff --git a/src/pkcs1_decode.c b/src/pkcs1_decode.c +index 207b198..74cb4a2 100644 +--- a/src/pkcs1_decode.c ++++ b/src/pkcs1_decode.c +@@ -130,7 +130,7 @@ STATIC size_t safe_select_idx(size_t in1, size_t in2, uint8_t choice) + * - in1[] is NOT equal to in2[] where neq_mask[] is 0xFF. + * Return non-zero otherwise. + */ +-STATIC uint8_t safe_cmp(const uint8_t *in1, const uint8_t *in2, ++STATIC uint8_t safe_cmp_masks(const uint8_t *in1, const uint8_t *in2, + const uint8_t *eq_mask, const uint8_t *neq_mask, + size_t len) + { +@@ -187,7 +187,7 @@ STATIC size_t safe_search(const uint8_t *in1, uint8_t c, size_t len) + return result; + } + +-#define EM_PREFIX_LEN 10 ++#define PKCS1_PREFIX_LEN 10 + + /* + * Decode and verify the PKCS#1 padding, then put either the plaintext +@@ -222,13 +222,13 @@ EXPORT_SYM int pkcs1_decode(const uint8_t *em, size_t len_em_output, + if (NULL == em || NULL == output || NULL == sentinel) { + return -1; + } +- if (len_em_output < (EM_PREFIX_LEN + 2)) { ++ if (len_em_output < (PKCS1_PREFIX_LEN + 2)) { + return -1; + } + if (len_sentinel > len_em_output) { + return -1; + } +- if (expected_pt_len > 0 && expected_pt_len > (len_em_output - EM_PREFIX_LEN - 1)) { ++ if (expected_pt_len > 0 && expected_pt_len > (len_em_output - PKCS1_PREFIX_LEN - 1)) { + return -1; + } + +@@ -240,7 +240,7 @@ EXPORT_SYM int pkcs1_decode(const uint8_t *em, size_t len_em_output, + memcpy(padded_sentinel + (len_em_output - len_sentinel), sentinel, len_sentinel); + + /** The first 10 bytes must follow the pattern **/ +- match = safe_cmp(em, ++ match = safe_cmp_masks(em, + (const uint8_t*)"\x00\x02" "\x00\x00\x00\x00\x00\x00\x00\x00", + (const uint8_t*)"\xFF\xFF" "\x00\x00\x00\x00\x00\x00\x00\x00", + (const uint8_t*)"\x00\x00" "\xFF\xFF\xFF\xFF\xFF\xFF\xFF\xFF", +@@ -283,3 +283,72 @@ end: + free(padded_sentinel); + return result; + } ++ ++/* ++ * Decode and verify the OAEP padding in constant time. ++ * ++ * The function returns the number of bytes to ignore at the beginning ++ * of db (the rest is the plaintext), or -1 in case of problems. ++ */ ++ ++EXPORT_SYM int oaep_decode(const uint8_t *em, ++ size_t em_len, ++ const uint8_t *lHash, ++ size_t hLen, ++ const uint8_t *db, ++ size_t db_len) /* em_len - 1 - hLen */ ++{ ++ int result; ++ size_t one_pos, search_len, i; ++ uint8_t wrong_padding; ++ uint8_t *eq_mask = NULL; ++ uint8_t *neq_mask = NULL; ++ uint8_t *target_db = NULL; ++ ++ if (NULL == em || NULL == lHash || NULL == db) { ++ return -1; ++ } ++ ++ if (em_len < 2*hLen+2 || db_len != em_len-1-hLen) { ++ return -1; ++ } ++ ++ /* Allocate */ ++ eq_mask = (uint8_t*) calloc(1, db_len); ++ neq_mask = (uint8_t*) calloc(1, db_len); ++ target_db = (uint8_t*) calloc(1, db_len); ++ if (NULL == eq_mask || NULL == neq_mask || NULL == target_db) { ++ result = -1; ++ goto cleanup; ++ } ++ ++ /* Step 3g */ ++ search_len = db_len - hLen; ++ ++ one_pos = safe_search(db + hLen, 0x01, search_len); ++ if (SIZE_T_MAX == one_pos) { ++ result = -1; ++ goto cleanup; ++ } ++ ++ memset(eq_mask, 0xAA, db_len); ++ memcpy(target_db, lHash, hLen); ++ memset(eq_mask, 0xFF, hLen); ++ ++ for (i=0; i<search_len; i++) { ++ eq_mask[hLen + i] = propagate_ones(i < one_pos); ++ } ++ ++ wrong_padding = em[0]; ++ wrong_padding |= safe_cmp_masks(db, target_db, eq_mask, neq_mask, db_len); ++ set_if_match(&wrong_padding, one_pos, search_len); ++ ++ result = wrong_padding ? -1 : (int)(hLen + 1 + one_pos); ++ ++cleanup: ++ free(eq_mask); ++ free(neq_mask); ++ free(target_db); ++ ++ return result; ++} +diff --git a/src/test/test_pkcs1.c b/src/test/test_pkcs1.c +index 6ef63cb..69aaac5 100644 +--- a/src/test/test_pkcs1.c ++++ b/src/test/test_pkcs1.c +@@ -5,7 +5,7 @@ void set_if_match(uint8_t *flag, size_t term1, size_t term2); + void set_if_no_match(uint8_t *flag, size_t term1, size_t term2); + void safe_select(const uint8_t *in1, const uint8_t *in2, uint8_t *out, uint8_t choice, size_t len); + size_t safe_select_idx(size_t in1, size_t in2, uint8_t choice); +-uint8_t safe_cmp(const uint8_t *in1, const uint8_t *in2, ++uint8_t safe_cmp_masks(const uint8_t *in1, const uint8_t *in2, + const uint8_t *eq_mask, const uint8_t *neq_mask, + size_t len); + size_t safe_search(const uint8_t *in1, uint8_t c, size_t len); +@@ -80,29 +80,29 @@ void test_safe_select_idx() + assert(safe_select_idx(0x100004, 0x223344, 1) == 0x223344); + } + +-void test_safe_cmp() ++void test_safe_cmp_masks(void) + { + uint8_t res; + +- res = safe_cmp(onezero, onezero, ++ res = safe_cmp_masks(onezero, onezero, + (uint8_t*)"\xFF\xFF", + (uint8_t*)"\x00\x00", + 2); + assert(res == 0); + +- res = safe_cmp(onezero, zerozero, ++ res = safe_cmp_masks(onezero, zerozero, + (uint8_t*)"\xFF\xFF", + (uint8_t*)"\x00\x00", + 2); + assert(res != 0); + +- res = safe_cmp(onezero, oneone, ++ res = safe_cmp_masks(onezero, oneone, + (uint8_t*)"\xFF\xFF", + (uint8_t*)"\x00\x00", + 2); + assert(res != 0); + +- res = safe_cmp(onezero, oneone, ++ res = safe_cmp_masks(onezero, oneone, + (uint8_t*)"\xFF\x00", + (uint8_t*)"\x00\x00", + 2); +@@ -110,19 +110,19 @@ void test_safe_cmp() + + /** -- **/ + +- res = safe_cmp(onezero, onezero, ++ res = safe_cmp_masks(onezero, onezero, + (uint8_t*)"\x00\x00", + (uint8_t*)"\xFF\xFF", + 2); + assert(res != 0); + +- res = safe_cmp(oneone, zerozero, ++ res = safe_cmp_masks(oneone, zerozero, + (uint8_t*)"\x00\x00", + (uint8_t*)"\xFF\xFF", + 2); + assert(res == 0); + +- res = safe_cmp(onezero, oneone, ++ res = safe_cmp_masks(onezero, oneone, + (uint8_t*)"\x00\x00", + (uint8_t*)"\x00\xFF", + 2); +@@ -130,7 +130,7 @@ void test_safe_cmp() + + /** -- **/ + +- res = safe_cmp(onezero, oneone, ++ res = safe_cmp_masks(onezero, oneone, + (uint8_t*)"\xFF\x00", + (uint8_t*)"\x00\xFF", + 2); +@@ -158,7 +158,7 @@ int main(void) + test_set_if_no_match(); + test_safe_select(); + test_safe_select_idx(); +- test_safe_cmp(); ++ test_safe_cmp_masks(); + test_safe_search(); + return 0; + } +-- +2.40.0 diff --git a/meta/recipes-devtools/python/python3-pycryptodome_3.14.1.bb b/meta/recipes-devtools/python/python3-pycryptodome_3.14.1.bb index c0324590c2..1e6c514224 100644 --- a/meta/recipes-devtools/python/python3-pycryptodome_3.14.1.bb +++ b/meta/recipes-devtools/python/python3-pycryptodome_3.14.1.bb @@ -3,3 +3,4 @@ inherit setuptools3 SRC_URI[sha256sum] = "e04e40a7f8c1669195536a37979dd87da2c32dbdc73d6fe35f0077b0c17c803b" +SRC_URI += "file://CVE-2023-52323.patch" diff --git a/meta/recipes-devtools/python/python3-pycryptodomex/CVE-2023-52323.patch b/meta/recipes-devtools/python/python3-pycryptodomex/CVE-2023-52323.patch new file mode 100644 index 0000000000..56000b996e --- /dev/null +++ b/meta/recipes-devtools/python/python3-pycryptodomex/CVE-2023-52323.patch @@ -0,0 +1,436 @@ +From 8ed5cf533be298d40ec9f75a188738ad4c3a8417 Mon Sep 17 00:00:00 2001 +From: Narpat Mali <narpat.mali@windriver.com> +Date: Thu, 8 Feb 2024 09:09:35 +0000 +Subject: [PATCH] Use constant-time (faster) padding decoding also for OAEP + +CVE: CVE-2023-52323 + +Upstream-Status: Backport [https://github.com/Legrandin/pycryptodome/commit/0deea1bfe1489e8c80d2053bbb06a1aa0b181ebd] + +Signed-off-by: Narpat Mali <narpat.mali@windriver.com> +--- + lib/Cryptodome/Cipher/PKCS1_OAEP.py | 38 +++++----- + lib/Cryptodome/Cipher/PKCS1_v1_5.py | 31 +------- + lib/Cryptodome/Cipher/_pkcs1_oaep_decode.py | 41 +++++++++++ + src/pkcs1_decode.c | 79 +++++++++++++++++++-- + src/test/test_pkcs1.c | 22 +++--- + 5 files changed, 145 insertions(+), 66 deletions(-) + create mode 100644 lib/Cryptodome/Cipher/_pkcs1_oaep_decode.py + +diff --git a/lib/Cryptodome/Cipher/PKCS1_OAEP.py b/lib/Cryptodome/Cipher/PKCS1_OAEP.py +index 7525c5d..653df04 100644 +--- a/lib/Cryptodome/Cipher/PKCS1_OAEP.py ++++ b/lib/Cryptodome/Cipher/PKCS1_OAEP.py +@@ -23,11 +23,13 @@ + from Cryptodome.Signature.pss import MGF1 + import Cryptodome.Hash.SHA1 + +-from Cryptodome.Util.py3compat import bord, _copy_bytes ++from Crypto.Util.py3compat import _copy_bytes + import Cryptodome.Util.number +-from Cryptodome.Util.number import ceil_div, bytes_to_long, long_to_bytes +-from Cryptodome.Util.strxor import strxor ++from Crypto.Util.number import ceil_div, bytes_to_long, long_to_bytes ++from Crypto.Util.strxor import strxor + from Cryptodome import Random ++from ._pkcs1_oaep_decode import oaep_decode ++ + + class PKCS1OAEP_Cipher: + """Cipher object for PKCS#1 v1.5 OAEP. +@@ -68,7 +70,7 @@ class PKCS1OAEP_Cipher: + if mgfunc: + self._mgf = mgfunc + else: +- self._mgf = lambda x,y: MGF1(x,y,self._hashObj) ++ self._mgf = lambda x, y: MGF1(x, y, self._hashObj) + + self._label = _copy_bytes(None, None, label) + self._randfunc = randfunc +@@ -105,7 +107,7 @@ class PKCS1OAEP_Cipher: + + # See 7.1.1 in RFC3447 + modBits = Cryptodome.Util.number.size(self._key.n) +- k = ceil_div(modBits, 8) # Convert from bits to bytes ++ k = ceil_div(modBits, 8) # Convert from bits to bytes + hLen = self._hashObj.digest_size + mLen = len(message) + +@@ -159,11 +161,11 @@ class PKCS1OAEP_Cipher: + + # See 7.1.2 in RFC3447 + modBits = Cryptodome.Util.number.size(self._key.n) +- k = ceil_div(modBits,8) # Convert from bits to bytes ++ k = ceil_div(modBits, 8) # Convert from bits to bytes + hLen = self._hashObj.digest_size + + # Step 1b and 1c +- if len(ciphertext) != k or k<hLen+2: ++ if len(ciphertext) != k or k < hLen+2: + raise ValueError("Ciphertext with incorrect length.") + # Step 2a (O2SIP) + ct_int = bytes_to_long(ciphertext) +@@ -173,8 +175,6 @@ class PKCS1OAEP_Cipher: + em = long_to_bytes(m_int, k) + # Step 3a + lHash = self._hashObj.new(self._label).digest() +- # Step 3b +- y = em[0] + # y must be 0, but we MUST NOT check it here in order not to + # allow attacks like Manger's (http://dl.acm.org/citation.cfm?id=704143) + maskedSeed = em[1:hLen+1] +@@ -187,22 +187,17 @@ class PKCS1OAEP_Cipher: + dbMask = self._mgf(seed, k-hLen-1) + # Step 3f + db = strxor(maskedDB, dbMask) +- # Step 3g +- one_pos = hLen + db[hLen:].find(b'\x01') +- lHash1 = db[:hLen] +- invalid = bord(y) | int(one_pos < hLen) +- hash_compare = strxor(lHash1, lHash) +- for x in hash_compare: +- invalid |= bord(x) +- for x in db[hLen:one_pos]: +- invalid |= bord(x) +- if invalid != 0: ++ # Step 3b + 3g ++ res = oaep_decode(em, lHash, db) ++ if res <= 0: + raise ValueError("Incorrect decryption.") + # Step 4 +- return db[one_pos + 1:] ++ return db[res:] ++ + + def new(key, hashAlgo=None, mgfunc=None, label=b'', randfunc=None): +- """Return a cipher object :class:`PKCS1OAEP_Cipher` that can be used to perform PKCS#1 OAEP encryption or decryption. ++ """Return a cipher object :class:`PKCS1OAEP_Cipher` ++ that can be used to perform PKCS#1 OAEP encryption or decryption. + + :param key: + The key object to use to encrypt or decrypt the message. +@@ -236,4 +231,3 @@ def new(key, hashAlgo=None, mgfunc=None, label=b'', randfunc=None): + if randfunc is None: + randfunc = Random.get_random_bytes + return PKCS1OAEP_Cipher(key, hashAlgo, mgfunc, label, randfunc) +- +diff --git a/lib/Cryptodome/Cipher/PKCS1_v1_5.py b/lib/Cryptodome/Cipher/PKCS1_v1_5.py +index 17ef9eb..f20a7ce 100644 +--- a/lib/Cryptodome/Cipher/PKCS1_v1_5.py ++++ b/lib/Cryptodome/Cipher/PKCS1_v1_5.py +@@ -25,31 +25,7 @@ __all__ = ['new', 'PKCS115_Cipher'] + from Cryptodome import Random + from Cryptodome.Util.number import bytes_to_long, long_to_bytes + from Cryptodome.Util.py3compat import bord, is_bytes, _copy_bytes +- +-from Cryptodome.Util._raw_api import (load_pycryptodome_raw_lib, c_size_t, +- c_uint8_ptr) +- +- +-_raw_pkcs1_decode = load_pycryptodome_raw_lib("Cryptodome.Cipher._pkcs1_decode", +- """ +- int pkcs1_decode(const uint8_t *em, size_t len_em, +- const uint8_t *sentinel, size_t len_sentinel, +- size_t expected_pt_len, +- uint8_t *output); +- """) +- +- +-def _pkcs1_decode(em, sentinel, expected_pt_len, output): +- if len(em) != len(output): +- raise ValueError("Incorrect output length") +- +- ret = _raw_pkcs1_decode.pkcs1_decode(c_uint8_ptr(em), +- c_size_t(len(em)), +- c_uint8_ptr(sentinel), +- c_size_t(len(sentinel)), +- c_size_t(expected_pt_len), +- c_uint8_ptr(output)) +- return ret ++from ._pkcs1_oaep_decode import pkcs1_decode + + + class PKCS115_Cipher: +@@ -113,7 +89,6 @@ class PKCS115_Cipher: + continue + ps.append(new_byte) + ps = b"".join(ps) +- assert(len(ps) == k - mLen - 3) + # Step 2b + em = b'\x00\x02' + ps + b'\x00' + _copy_bytes(None, None, message) + # Step 3a (OS2IP) +@@ -185,14 +160,14 @@ class PKCS115_Cipher: + # Step 3 (not constant time when the sentinel is not a byte string) + output = bytes(bytearray(k)) + if not is_bytes(sentinel) or len(sentinel) > k: +- size = _pkcs1_decode(em, b'', expected_pt_len, output) ++ size = pkcs1_decode(em, b'', expected_pt_len, output) + if size < 0: + return sentinel + else: + return output[size:] + + # Step 3 (somewhat constant time) +- size = _pkcs1_decode(em, sentinel, expected_pt_len, output) ++ size = pkcs1_decode(em, sentinel, expected_pt_len, output) + return output[size:] + + +diff --git a/lib/Cryptodome/Cipher/_pkcs1_oaep_decode.py b/lib/Cryptodome/Cipher/_pkcs1_oaep_decode.py +new file mode 100644 +index 0000000..fc07528 +--- /dev/null ++++ b/lib/Cryptodome/Cipher/_pkcs1_oaep_decode.py +@@ -0,0 +1,41 @@ ++from Crypto.Util._raw_api import (load_pycryptodome_raw_lib, c_size_t, ++ c_uint8_ptr) ++ ++ ++_raw_pkcs1_decode = load_pycryptodome_raw_lib("Crypto.Cipher._pkcs1_decode", ++ """ ++ int pkcs1_decode(const uint8_t *em, size_t len_em, ++ const uint8_t *sentinel, size_t len_sentinel, ++ size_t expected_pt_len, ++ uint8_t *output); ++ ++ int oaep_decode(const uint8_t *em, ++ size_t em_len, ++ const uint8_t *lHash, ++ size_t hLen, ++ const uint8_t *db, ++ size_t db_len); ++ """) ++ ++ ++def pkcs1_decode(em, sentinel, expected_pt_len, output): ++ if len(em) != len(output): ++ raise ValueError("Incorrect output length") ++ ++ ret = _raw_pkcs1_decode.pkcs1_decode(c_uint8_ptr(em), ++ c_size_t(len(em)), ++ c_uint8_ptr(sentinel), ++ c_size_t(len(sentinel)), ++ c_size_t(expected_pt_len), ++ c_uint8_ptr(output)) ++ return ret ++ ++ ++def oaep_decode(em, lHash, db): ++ ret = _raw_pkcs1_decode.oaep_decode(c_uint8_ptr(em), ++ c_size_t(len(em)), ++ c_uint8_ptr(lHash), ++ c_size_t(len(lHash)), ++ c_uint8_ptr(db), ++ c_size_t(len(db))) ++ return ret +diff --git a/src/pkcs1_decode.c b/src/pkcs1_decode.c +index 207b198..74cb4a2 100644 +--- a/src/pkcs1_decode.c ++++ b/src/pkcs1_decode.c +@@ -130,7 +130,7 @@ STATIC size_t safe_select_idx(size_t in1, size_t in2, uint8_t choice) + * - in1[] is NOT equal to in2[] where neq_mask[] is 0xFF. + * Return non-zero otherwise. + */ +-STATIC uint8_t safe_cmp(const uint8_t *in1, const uint8_t *in2, ++STATIC uint8_t safe_cmp_masks(const uint8_t *in1, const uint8_t *in2, + const uint8_t *eq_mask, const uint8_t *neq_mask, + size_t len) + { +@@ -187,7 +187,7 @@ STATIC size_t safe_search(const uint8_t *in1, uint8_t c, size_t len) + return result; + } + +-#define EM_PREFIX_LEN 10 ++#define PKCS1_PREFIX_LEN 10 + + /* + * Decode and verify the PKCS#1 padding, then put either the plaintext +@@ -222,13 +222,13 @@ EXPORT_SYM int pkcs1_decode(const uint8_t *em, size_t len_em_output, + if (NULL == em || NULL == output || NULL == sentinel) { + return -1; + } +- if (len_em_output < (EM_PREFIX_LEN + 2)) { ++ if (len_em_output < (PKCS1_PREFIX_LEN + 2)) { + return -1; + } + if (len_sentinel > len_em_output) { + return -1; + } +- if (expected_pt_len > 0 && expected_pt_len > (len_em_output - EM_PREFIX_LEN - 1)) { ++ if (expected_pt_len > 0 && expected_pt_len > (len_em_output - PKCS1_PREFIX_LEN - 1)) { + return -1; + } + +@@ -240,7 +240,7 @@ EXPORT_SYM int pkcs1_decode(const uint8_t *em, size_t len_em_output, + memcpy(padded_sentinel + (len_em_output - len_sentinel), sentinel, len_sentinel); + + /** The first 10 bytes must follow the pattern **/ +- match = safe_cmp(em, ++ match = safe_cmp_masks(em, + (const uint8_t*)"\x00\x02" "\x00\x00\x00\x00\x00\x00\x00\x00", + (const uint8_t*)"\xFF\xFF" "\x00\x00\x00\x00\x00\x00\x00\x00", + (const uint8_t*)"\x00\x00" "\xFF\xFF\xFF\xFF\xFF\xFF\xFF\xFF", +@@ -283,3 +283,72 @@ end: + free(padded_sentinel); + return result; + } ++ ++/* ++ * Decode and verify the OAEP padding in constant time. ++ * ++ * The function returns the number of bytes to ignore at the beginning ++ * of db (the rest is the plaintext), or -1 in case of problems. ++ */ ++ ++EXPORT_SYM int oaep_decode(const uint8_t *em, ++ size_t em_len, ++ const uint8_t *lHash, ++ size_t hLen, ++ const uint8_t *db, ++ size_t db_len) /* em_len - 1 - hLen */ ++{ ++ int result; ++ size_t one_pos, search_len, i; ++ uint8_t wrong_padding; ++ uint8_t *eq_mask = NULL; ++ uint8_t *neq_mask = NULL; ++ uint8_t *target_db = NULL; ++ ++ if (NULL == em || NULL == lHash || NULL == db) { ++ return -1; ++ } ++ ++ if (em_len < 2*hLen+2 || db_len != em_len-1-hLen) { ++ return -1; ++ } ++ ++ /* Allocate */ ++ eq_mask = (uint8_t*) calloc(1, db_len); ++ neq_mask = (uint8_t*) calloc(1, db_len); ++ target_db = (uint8_t*) calloc(1, db_len); ++ if (NULL == eq_mask || NULL == neq_mask || NULL == target_db) { ++ result = -1; ++ goto cleanup; ++ } ++ ++ /* Step 3g */ ++ search_len = db_len - hLen; ++ ++ one_pos = safe_search(db + hLen, 0x01, search_len); ++ if (SIZE_T_MAX == one_pos) { ++ result = -1; ++ goto cleanup; ++ } ++ ++ memset(eq_mask, 0xAA, db_len); ++ memcpy(target_db, lHash, hLen); ++ memset(eq_mask, 0xFF, hLen); ++ ++ for (i=0; i<search_len; i++) { ++ eq_mask[hLen + i] = propagate_ones(i < one_pos); ++ } ++ ++ wrong_padding = em[0]; ++ wrong_padding |= safe_cmp_masks(db, target_db, eq_mask, neq_mask, db_len); ++ set_if_match(&wrong_padding, one_pos, search_len); ++ ++ result = wrong_padding ? -1 : (int)(hLen + 1 + one_pos); ++ ++cleanup: ++ free(eq_mask); ++ free(neq_mask); ++ free(target_db); ++ ++ return result; ++} +diff --git a/src/test/test_pkcs1.c b/src/test/test_pkcs1.c +index 6ef63cb..69aaac5 100644 +--- a/src/test/test_pkcs1.c ++++ b/src/test/test_pkcs1.c +@@ -5,7 +5,7 @@ void set_if_match(uint8_t *flag, size_t term1, size_t term2); + void set_if_no_match(uint8_t *flag, size_t term1, size_t term2); + void safe_select(const uint8_t *in1, const uint8_t *in2, uint8_t *out, uint8_t choice, size_t len); + size_t safe_select_idx(size_t in1, size_t in2, uint8_t choice); +-uint8_t safe_cmp(const uint8_t *in1, const uint8_t *in2, ++uint8_t safe_cmp_masks(const uint8_t *in1, const uint8_t *in2, + const uint8_t *eq_mask, const uint8_t *neq_mask, + size_t len); + size_t safe_search(const uint8_t *in1, uint8_t c, size_t len); +@@ -80,29 +80,29 @@ void test_safe_select_idx() + assert(safe_select_idx(0x100004, 0x223344, 1) == 0x223344); + } + +-void test_safe_cmp() ++void test_safe_cmp_masks(void) + { + uint8_t res; + +- res = safe_cmp(onezero, onezero, ++ res = safe_cmp_masks(onezero, onezero, + (uint8_t*)"\xFF\xFF", + (uint8_t*)"\x00\x00", + 2); + assert(res == 0); + +- res = safe_cmp(onezero, zerozero, ++ res = safe_cmp_masks(onezero, zerozero, + (uint8_t*)"\xFF\xFF", + (uint8_t*)"\x00\x00", + 2); + assert(res != 0); + +- res = safe_cmp(onezero, oneone, ++ res = safe_cmp_masks(onezero, oneone, + (uint8_t*)"\xFF\xFF", + (uint8_t*)"\x00\x00", + 2); + assert(res != 0); + +- res = safe_cmp(onezero, oneone, ++ res = safe_cmp_masks(onezero, oneone, + (uint8_t*)"\xFF\x00", + (uint8_t*)"\x00\x00", + 2); +@@ -110,19 +110,19 @@ void test_safe_cmp() + + /** -- **/ + +- res = safe_cmp(onezero, onezero, ++ res = safe_cmp_masks(onezero, onezero, + (uint8_t*)"\x00\x00", + (uint8_t*)"\xFF\xFF", + 2); + assert(res != 0); + +- res = safe_cmp(oneone, zerozero, ++ res = safe_cmp_masks(oneone, zerozero, + (uint8_t*)"\x00\x00", + (uint8_t*)"\xFF\xFF", + 2); + assert(res == 0); + +- res = safe_cmp(onezero, oneone, ++ res = safe_cmp_masks(onezero, oneone, + (uint8_t*)"\x00\x00", + (uint8_t*)"\x00\xFF", + 2); +@@ -130,7 +130,7 @@ void test_safe_cmp() + + /** -- **/ + +- res = safe_cmp(onezero, oneone, ++ res = safe_cmp_masks(onezero, oneone, + (uint8_t*)"\xFF\x00", + (uint8_t*)"\x00\xFF", + 2); +@@ -158,7 +158,7 @@ int main(void) + test_set_if_no_match(); + test_safe_select(); + test_safe_select_idx(); +- test_safe_cmp(); ++ test_safe_cmp_masks(); + test_safe_search(); + return 0; + } +-- +2.40.0 diff --git a/meta/recipes-devtools/python/python3-pycryptodomex_3.14.1.bb b/meta/recipes-devtools/python/python3-pycryptodomex_3.14.1.bb index 79a3fee19c..31ad3fda5e 100644 --- a/meta/recipes-devtools/python/python3-pycryptodomex_3.14.1.bb +++ b/meta/recipes-devtools/python/python3-pycryptodomex_3.14.1.bb @@ -3,6 +3,8 @@ inherit setuptools3 SRC_URI[sha256sum] = "2ce76ed0081fd6ac8c74edc75b9d14eca2064173af79843c24fa62573263c1f2" +SRC_URI += "file://CVE-2023-52323.patch" + FILES:${PN}-tests = " \ ${PYTHON_SITEPACKAGES_DIR}/Cryptodome/SelfTest/ \ ${PYTHON_SITEPACKAGES_DIR}/Cryptodome/SelfTest/__pycache__/ \ -- 2.34.1 ^ permalink raw reply related [flat|nested] 33+ messages in thread
* [OE-core][kirkstone 6/8] systemd: Only add myhostname to nsswitch.conf if in PACKAGECONFIG 2024-02-12 13:54 [OE-core][kirkstone 0/8] Patch review Steve Sakoman ` (4 preceding siblings ...) 2024-02-12 13:54 ` [OE-core][kirkstone 5/8] python3-pycryptodome: Fix CVE-2023-52323 Steve Sakoman @ 2024-02-12 13:54 ` Steve Sakoman 2024-02-12 13:54 ` [OE-core][kirkstone 7/8] kernel: fix localversion in v6.3+ Steve Sakoman 2024-02-12 13:54 ` [OE-core][kirkstone 8/8] ghostscript: correct LICENSE with AGPLv3 Steve Sakoman 7 siblings, 0 replies; 33+ messages in thread From: Steve Sakoman @ 2024-02-12 13:54 UTC (permalink / raw) To: openembedded-core From: Jermain Horsman <jermain.horsman@nedap.com> Currently myhostname is always added to nsswitch.conf even if it is not included in PACKAGECONFIG. This is based on changes made in OE-core rev: ba3a78c08cb0ce08afde049610d3172b9e3b0695 Cc: Chen Qi <Qi.Chen@windriver.com> Signed-off-by: Jermain Horsman <jermain.horsman@nedap.com> Signed-off-by: Steve Sakoman <steve@sakoman.com> --- meta/recipes-core/systemd/systemd_250.5.bb | 16 ++++++++++------ 1 file changed, 10 insertions(+), 6 deletions(-) diff --git a/meta/recipes-core/systemd/systemd_250.5.bb b/meta/recipes-core/systemd/systemd_250.5.bb index 889473ee1f..80a797cf2c 100644 --- a/meta/recipes-core/systemd/systemd_250.5.bb +++ b/meta/recipes-core/systemd/systemd_250.5.bb @@ -776,15 +776,19 @@ ALTERNATIVE_LINK_NAME[runlevel] = "${base_sbindir}/runlevel" ALTERNATIVE_PRIORITY[runlevel] ?= "300" pkg_postinst:${PN}:libc-glibc () { - sed -e '/^hosts:/s/\s*\<myhostname\>//' \ - -e 's/\(^hosts:.*\)\(\<files\>\)\(.*\)\(\<dns\>\)\(.*\)/\1\2 myhostname \3\4\5/' \ - -i $D${sysconfdir}/nsswitch.conf + if ${@bb.utils.contains('PACKAGECONFIG', 'myhostname', 'true', 'false', d)}; then + sed -e '/^hosts:/s/\s*\<myhostname\>//' \ + -e 's/\(^hosts:.*\)\(\<files\>\)\(.*\)\(\<dns\>\)\(.*\)/\1\2 myhostname \3\4\5/' \ + -i $D${sysconfdir}/nsswitch.conf + fi } pkg_prerm:${PN}:libc-glibc () { - sed -e '/^hosts:/s/\s*\<myhostname\>//' \ - -e '/^hosts:/s/\s*myhostname//' \ - -i $D${sysconfdir}/nsswitch.conf + if ${@bb.utils.contains('PACKAGECONFIG', 'myhostname', 'true', 'false', d)}; then + sed -e '/^hosts:/s/\s*\<myhostname\>//' \ + -e '/^hosts:/s/\s*myhostname//' \ + -i $D${sysconfdir}/nsswitch.conf + fi } PACKAGE_WRITE_DEPS += "qemu-native" -- 2.34.1 ^ permalink raw reply related [flat|nested] 33+ messages in thread
* [OE-core][kirkstone 7/8] kernel: fix localversion in v6.3+ 2024-02-12 13:54 [OE-core][kirkstone 0/8] Patch review Steve Sakoman ` (5 preceding siblings ...) 2024-02-12 13:54 ` [OE-core][kirkstone 6/8] systemd: Only add myhostname to nsswitch.conf if in PACKAGECONFIG Steve Sakoman @ 2024-02-12 13:54 ` Steve Sakoman 2024-02-16 1:20 ` Ryan Eatmon 2024-02-12 13:54 ` [OE-core][kirkstone 8/8] ghostscript: correct LICENSE with AGPLv3 Steve Sakoman 7 siblings, 1 reply; 33+ messages in thread From: Steve Sakoman @ 2024-02-12 13:54 UTC (permalink / raw) To: openembedded-core From: Bruce Ashfield <bruce.ashfield@gmail.com> During testing of the v6.4 reference kernel, it was noticed that on-target modules no longer matched the magic value of the running kernel. This was due to a different localversion in the cross built kernel and the scripts / resources created on target. This was due to changes in the setlocalversion script introduced in the v6.3 series. The .scmversion file is no longer used (or packaged) to inhibit the addition of a "+" (through querying of the git status of the kernel) or the setting of a local version. We recently introduced the KERNEL_LOCALVERSION variable to allow recipes to place a value in .scmversion, so we extend the use of that variable to kernel-arch.bbclass and use it to set the exported variable LOCALVERSION. We must do it at the kernel-arch level, as the variable must be exported in any kernel build to ensure that setlocalversion always correctly sets the localversion. Signed-off-by: Bruce Ashfield <bruce.ashfield@gmail.com> Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org> Cherry-picked from master 765b13b7305c8d2f222cfc66d77c02e6a088c691 Signed-off-by: Andreas Helbech Kleist <andreaskleist@gmail.com> Signed-off-by: Steve Sakoman <steve@sakoman.com> --- meta/classes/kernel-arch.bbclass | 7 +++++++ meta/classes/kernel.bbclass | 10 ++++++++-- 2 files changed, 15 insertions(+), 2 deletions(-) diff --git a/meta/classes/kernel-arch.bbclass b/meta/classes/kernel-arch.bbclass index 4cd08b96fb..0a79dea0af 100644 --- a/meta/classes/kernel-arch.bbclass +++ b/meta/classes/kernel-arch.bbclass @@ -66,3 +66,10 @@ KERNEL_LD = "${CCACHE}${HOST_PREFIX}ld.bfd ${HOST_LD_KERNEL_ARCH}" KERNEL_AR = "${CCACHE}${HOST_PREFIX}ar ${HOST_AR_KERNEL_ARCH}" TOOLCHAIN ?= "gcc" +# 6.3+ requires the variable LOCALVERSION to be set to not get a "+" in +# the local version. Having it empty means nothing will be added, and any +# value will be appended to the local kernel version. This replaces the +# use of .scmversion file for setting a localversion without using +# the CONFIG_LOCALVERSION option. +KERNEL_LOCALVERSION ??= "" +export LOCALVERSION ?= "${KERNEL_LOCALVERSION}" diff --git a/meta/classes/kernel.bbclass b/meta/classes/kernel.bbclass index 5951347361..940f1a3cf4 100644 --- a/meta/classes/kernel.bbclass +++ b/meta/classes/kernel.bbclass @@ -418,7 +418,7 @@ do_compile_kernelmodules() { if (grep -q -i -e '^CONFIG_MODULES=y$' ${B}/.config); then oe_runmake -C ${B} ${PARALLEL_MAKE} modules ${KERNEL_EXTRA_ARGS} - # Module.symvers gets updated during the + # Module.symvers gets updated during the # building of the kernel modules. We need to # update this in the shared workdir since some # external kernel modules has a dependency on @@ -635,7 +635,13 @@ kernel_do_configure() { # $ scripts/setlocalversion . => + # $ make kernelversion => 2.6.37 # $ make kernelrelease => 2.6.37+ - touch ${B}/.scmversion ${S}/.scmversion + # See kernel-arch.bbclass for post v6.3 removal of the extra + # + in localversion. .scmversion is no longer used, and the + # variable LOCALVERSION must be used + if [ ! -e ${B}/.scmversion -a ! -e ${S}/.scmversion ]; then + echo ${KERNEL_LOCALVERSION} > ${B}/.scmversion + echo ${KERNEL_LOCALVERSION} > ${S}/.scmversion + fi if [ "${S}" != "${B}" ] && [ -f "${S}/.config" ] && [ ! -f "${B}/.config" ]; then mv "${S}/.config" "${B}/.config" -- 2.34.1 ^ permalink raw reply related [flat|nested] 33+ messages in thread
* Re: [OE-core][kirkstone 7/8] kernel: fix localversion in v6.3+ 2024-02-12 13:54 ` [OE-core][kirkstone 7/8] kernel: fix localversion in v6.3+ Steve Sakoman @ 2024-02-16 1:20 ` Ryan Eatmon 2024-02-16 1:43 ` Steve Sakoman 0 siblings, 1 reply; 33+ messages in thread From: Ryan Eatmon @ 2024-02-16 1:20 UTC (permalink / raw) To: Steve Sakoman, openembedded-core With this patch in place we are seeing a breakage on our kernel builds. This patch was cherry picked from master, but the next patch on this file in master removes the LOCALVERSION setting... https://git.openembedded.org/openembedded-core/commit/meta/classes-recipe/kernel-arch.bbclass?h=master-next&id=b378eec156998eea55ba61e59103cb34fab0d07c There is a disconnect here. Why did we change kirkstone, a stable LTS version with a partial patch series based on master? On 2/12/2024 7:54 AM, Steve Sakoman wrote: > From: Bruce Ashfield <bruce.ashfield@gmail.com> > > During testing of the v6.4 reference kernel, it was noticed that > on-target modules no longer matched the magic value of the running > kernel. > > This was due to a different localversion in the cross built kernel > and the scripts / resources created on target. > > This was due to changes in the setlocalversion script introduced > in the v6.3 series. > > The .scmversion file is no longer used (or packaged) to inhibit > the addition of a "+" (through querying of the git status of the > kernel) or the setting of a local version. > > We recently introduced the KERNEL_LOCALVERSION variable to allow > recipes to place a value in .scmversion, so we extend the use of > that variable to kernel-arch.bbclass and use it to set the > exported variable LOCALVERSION. > > We must do it at the kernel-arch level, as the variable must be > exported in any kernel build to ensure that setlocalversion always > correctly sets the localversion. > > Signed-off-by: Bruce Ashfield <bruce.ashfield@gmail.com> > Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org> > > Cherry-picked from master 765b13b7305c8d2f222cfc66d77c02e6a088c691 > > Signed-off-by: Andreas Helbech Kleist <andreaskleist@gmail.com> > Signed-off-by: Steve Sakoman <steve@sakoman.com> > --- > meta/classes/kernel-arch.bbclass | 7 +++++++ > meta/classes/kernel.bbclass | 10 ++++++++-- > 2 files changed, 15 insertions(+), 2 deletions(-) > > diff --git a/meta/classes/kernel-arch.bbclass b/meta/classes/kernel-arch.bbclass > index 4cd08b96fb..0a79dea0af 100644 > --- a/meta/classes/kernel-arch.bbclass > +++ b/meta/classes/kernel-arch.bbclass > @@ -66,3 +66,10 @@ KERNEL_LD = "${CCACHE}${HOST_PREFIX}ld.bfd ${HOST_LD_KERNEL_ARCH}" > KERNEL_AR = "${CCACHE}${HOST_PREFIX}ar ${HOST_AR_KERNEL_ARCH}" > TOOLCHAIN ?= "gcc" > > +# 6.3+ requires the variable LOCALVERSION to be set to not get a "+" in > +# the local version. Having it empty means nothing will be added, and any > +# value will be appended to the local kernel version. This replaces the > +# use of .scmversion file for setting a localversion without using > +# the CONFIG_LOCALVERSION option. > +KERNEL_LOCALVERSION ??= "" > +export LOCALVERSION ?= "${KERNEL_LOCALVERSION}" > diff --git a/meta/classes/kernel.bbclass b/meta/classes/kernel.bbclass > index 5951347361..940f1a3cf4 100644 > --- a/meta/classes/kernel.bbclass > +++ b/meta/classes/kernel.bbclass > @@ -418,7 +418,7 @@ do_compile_kernelmodules() { > if (grep -q -i -e '^CONFIG_MODULES=y$' ${B}/.config); then > oe_runmake -C ${B} ${PARALLEL_MAKE} modules ${KERNEL_EXTRA_ARGS} > > - # Module.symvers gets updated during the > + # Module.symvers gets updated during the > # building of the kernel modules. We need to > # update this in the shared workdir since some > # external kernel modules has a dependency on > @@ -635,7 +635,13 @@ kernel_do_configure() { > # $ scripts/setlocalversion . => + > # $ make kernelversion => 2.6.37 > # $ make kernelrelease => 2.6.37+ > - touch ${B}/.scmversion ${S}/.scmversion > + # See kernel-arch.bbclass for post v6.3 removal of the extra > + # + in localversion. .scmversion is no longer used, and the > + # variable LOCALVERSION must be used > + if [ ! -e ${B}/.scmversion -a ! -e ${S}/.scmversion ]; then > + echo ${KERNEL_LOCALVERSION} > ${B}/.scmversion > + echo ${KERNEL_LOCALVERSION} > ${S}/.scmversion > + fi > > if [ "${S}" != "${B}" ] && [ -f "${S}/.config" ] && [ ! -f "${B}/.config" ]; then > mv "${S}/.config" "${B}/.config" > > > > -=-=-=-=-=-=-=-=-=-=-=- > Links: You receive all messages sent to this group. > View/Reply Online (#195328): https://lists.openembedded.org/g/openembedded-core/message/195328 > Mute This Topic: https://lists.openembedded.org/mt/104311283/6551054 > Group Owner: openembedded-core+owner@lists.openembedded.org > Unsubscribe: https://lists.openembedded.org/g/openembedded-core/unsub [reatmon@ti.com] > -=-=-=-=-=-=-=-=-=-=-=- > -- Ryan Eatmon reatmon@ti.com ----------------------------------------- Texas Instruments, Inc. - LCPD - MGTS ^ permalink raw reply [flat|nested] 33+ messages in thread
* Re: [OE-core][kirkstone 7/8] kernel: fix localversion in v6.3+ 2024-02-16 1:20 ` Ryan Eatmon @ 2024-02-16 1:43 ` Steve Sakoman 2024-02-16 1:45 ` Ryan Eatmon 0 siblings, 1 reply; 33+ messages in thread From: Steve Sakoman @ 2024-02-16 1:43 UTC (permalink / raw) To: Ryan Eatmon; +Cc: openembedded-core On Thu, Feb 15, 2024 at 3:20 PM Ryan Eatmon <reatmon@ti.com> wrote: > > > With this patch in place we are seeing a breakage on our kernel builds. > This patch was cherry picked from master, but the next patch on this > file in master removes the LOCALVERSION setting... > > https://git.openembedded.org/openembedded-core/commit/meta/classes-recipe/kernel-arch.bbclass?h=master-next&id=b378eec156998eea55ba61e59103cb34fab0d07c > > > There is a disconnect here. Why did we change kirkstone, a stable LTS > version with a partial patch series based on master? The original kirkstone backport request for this patch was sent to the list for review on February 9. There were no comments or objections, so I added it to the patch test queue. No issues were encountered on the autobuilder, so I sent this patch (along with the rest of the patch queue) to the list for a second review opportunity on February 12. Once again there were no comments or objections, so it was merged today February 15. I follow this process because I'm not smart enough, nor do I have time enough to thoroughly research all implications of every patch. I rely on autobuilder testing and community review to minimize breakage. This normally works quite well, but this time it didn't. I'll revert this patch and if someone would like to resubmit a proper series to deal with the issue it is trying to fix I will consider it and run it through the same process. Steve > On 2/12/2024 7:54 AM, Steve Sakoman wrote: > > From: Bruce Ashfield <bruce.ashfield@gmail.com> > > > > During testing of the v6.4 reference kernel, it was noticed that > > on-target modules no longer matched the magic value of the running > > kernel. > > > > This was due to a different localversion in the cross built kernel > > and the scripts / resources created on target. > > > > This was due to changes in the setlocalversion script introduced > > in the v6.3 series. > > > > The .scmversion file is no longer used (or packaged) to inhibit > > the addition of a "+" (through querying of the git status of the > > kernel) or the setting of a local version. > > > > We recently introduced the KERNEL_LOCALVERSION variable to allow > > recipes to place a value in .scmversion, so we extend the use of > > that variable to kernel-arch.bbclass and use it to set the > > exported variable LOCALVERSION. > > > > We must do it at the kernel-arch level, as the variable must be > > exported in any kernel build to ensure that setlocalversion always > > correctly sets the localversion. > > > > Signed-off-by: Bruce Ashfield <bruce.ashfield@gmail.com> > > Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org> > > > > Cherry-picked from master 765b13b7305c8d2f222cfc66d77c02e6a088c691 > > > > Signed-off-by: Andreas Helbech Kleist <andreaskleist@gmail.com> > > Signed-off-by: Steve Sakoman <steve@sakoman.com> > > --- > > meta/classes/kernel-arch.bbclass | 7 +++++++ > > meta/classes/kernel.bbclass | 10 ++++++++-- > > 2 files changed, 15 insertions(+), 2 deletions(-) > > > > diff --git a/meta/classes/kernel-arch.bbclass b/meta/classes/kernel-arch.bbclass > > index 4cd08b96fb..0a79dea0af 100644 > > --- a/meta/classes/kernel-arch.bbclass > > +++ b/meta/classes/kernel-arch.bbclass > > @@ -66,3 +66,10 @@ KERNEL_LD = "${CCACHE}${HOST_PREFIX}ld.bfd ${HOST_LD_KERNEL_ARCH}" > > KERNEL_AR = "${CCACHE}${HOST_PREFIX}ar ${HOST_AR_KERNEL_ARCH}" > > TOOLCHAIN ?= "gcc" > > > > +# 6.3+ requires the variable LOCALVERSION to be set to not get a "+" in > > +# the local version. Having it empty means nothing will be added, and any > > +# value will be appended to the local kernel version. This replaces the > > +# use of .scmversion file for setting a localversion without using > > +# the CONFIG_LOCALVERSION option. > > +KERNEL_LOCALVERSION ??= "" > > +export LOCALVERSION ?= "${KERNEL_LOCALVERSION}" > > diff --git a/meta/classes/kernel.bbclass b/meta/classes/kernel.bbclass > > index 5951347361..940f1a3cf4 100644 > > --- a/meta/classes/kernel.bbclass > > +++ b/meta/classes/kernel.bbclass > > @@ -418,7 +418,7 @@ do_compile_kernelmodules() { > > if (grep -q -i -e '^CONFIG_MODULES=y$' ${B}/.config); then > > oe_runmake -C ${B} ${PARALLEL_MAKE} modules ${KERNEL_EXTRA_ARGS} > > > > - # Module.symvers gets updated during the > > + # Module.symvers gets updated during the > > # building of the kernel modules. We need to > > # update this in the shared workdir since some > > # external kernel modules has a dependency on > > @@ -635,7 +635,13 @@ kernel_do_configure() { > > # $ scripts/setlocalversion . => + > > # $ make kernelversion => 2.6.37 > > # $ make kernelrelease => 2.6.37+ > > - touch ${B}/.scmversion ${S}/.scmversion > > + # See kernel-arch.bbclass for post v6.3 removal of the extra > > + # + in localversion. .scmversion is no longer used, and the > > + # variable LOCALVERSION must be used > > + if [ ! -e ${B}/.scmversion -a ! -e ${S}/.scmversion ]; then > > + echo ${KERNEL_LOCALVERSION} > ${B}/.scmversion > > + echo ${KERNEL_LOCALVERSION} > ${S}/.scmversion > > + fi > > > > if [ "${S}" != "${B}" ] && [ -f "${S}/.config" ] && [ ! -f "${B}/.config" ]; then > > mv "${S}/.config" "${B}/.config" > > > > > > > > -=-=-=-=-=-=-=-=-=-=-=- > > Links: You receive all messages sent to this group. > > View/Reply Online (#195328): https://lists.openembedded.org/g/openembedded-core/message/195328 > > Mute This Topic: https://lists.openembedded.org/mt/104311283/6551054 > > Group Owner: openembedded-core+owner@lists.openembedded.org > > Unsubscribe: https://lists.openembedded.org/g/openembedded-core/unsub [reatmon@ti.com] > > -=-=-=-=-=-=-=-=-=-=-=- > > > > -- > Ryan Eatmon reatmon@ti.com > ----------------------------------------- > Texas Instruments, Inc. - LCPD - MGTS ^ permalink raw reply [flat|nested] 33+ messages in thread
* Re: [OE-core][kirkstone 7/8] kernel: fix localversion in v6.3+ 2024-02-16 1:43 ` Steve Sakoman @ 2024-02-16 1:45 ` Ryan Eatmon 2024-02-16 8:11 ` Andreas Helbech Kleist 0 siblings, 1 reply; 33+ messages in thread From: Ryan Eatmon @ 2024-02-16 1:45 UTC (permalink / raw) To: Steve Sakoman; +Cc: openembedded-core On 2/15/2024 7:43 PM, Steve Sakoman wrote: > On Thu, Feb 15, 2024 at 3:20 PM Ryan Eatmon <reatmon@ti.com> wrote: >> >> >> With this patch in place we are seeing a breakage on our kernel builds. >> This patch was cherry picked from master, but the next patch on this >> file in master removes the LOCALVERSION setting... >> >> https://git.openembedded.org/openembedded-core/commit/meta/classes-recipe/kernel-arch.bbclass?h=master-next&id=b378eec156998eea55ba61e59103cb34fab0d07c >> >> >> There is a disconnect here. Why did we change kirkstone, a stable LTS >> version with a partial patch series based on master? > > The original kirkstone backport request for this patch was sent to the > list for review on February 9. > > There were no comments or objections, so I added it to the patch test > queue. No issues were encountered on the autobuilder, so I sent this > patch (along with the rest of the patch queue) to the list for a > second review opportunity on February 12. Once again there were no > comments or objections, so it was merged today February 15. > > I follow this process because I'm not smart enough, nor do I have time > enough to thoroughly research all implications of every patch. I rely > on autobuilder testing and community review to minimize breakage. > > This normally works quite well, but this time it didn't. I'll revert > this patch and if someone would like to resubmit a proper series to > deal with the issue it is trying to fix I will consider it and run it > through the same process. I can appreciate that. My comment was more for the person who sent the patch in the first place. You are doing a great job. I will try and pay attention for a replacement patch coming in and review as well. > Steve > > >> On 2/12/2024 7:54 AM, Steve Sakoman wrote: >>> From: Bruce Ashfield <bruce.ashfield@gmail.com> >>> >>> During testing of the v6.4 reference kernel, it was noticed that >>> on-target modules no longer matched the magic value of the running >>> kernel. >>> >>> This was due to a different localversion in the cross built kernel >>> and the scripts / resources created on target. >>> >>> This was due to changes in the setlocalversion script introduced >>> in the v6.3 series. >>> >>> The .scmversion file is no longer used (or packaged) to inhibit >>> the addition of a "+" (through querying of the git status of the >>> kernel) or the setting of a local version. >>> >>> We recently introduced the KERNEL_LOCALVERSION variable to allow >>> recipes to place a value in .scmversion, so we extend the use of >>> that variable to kernel-arch.bbclass and use it to set the >>> exported variable LOCALVERSION. >>> >>> We must do it at the kernel-arch level, as the variable must be >>> exported in any kernel build to ensure that setlocalversion always >>> correctly sets the localversion. >>> >>> Signed-off-by: Bruce Ashfield <bruce.ashfield@gmail.com> >>> Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org> >>> >>> Cherry-picked from master 765b13b7305c8d2f222cfc66d77c02e6a088c691 >>> >>> Signed-off-by: Andreas Helbech Kleist <andreaskleist@gmail.com> >>> Signed-off-by: Steve Sakoman <steve@sakoman.com> >>> --- >>> meta/classes/kernel-arch.bbclass | 7 +++++++ >>> meta/classes/kernel.bbclass | 10 ++++++++-- >>> 2 files changed, 15 insertions(+), 2 deletions(-) >>> >>> diff --git a/meta/classes/kernel-arch.bbclass b/meta/classes/kernel-arch.bbclass >>> index 4cd08b96fb..0a79dea0af 100644 >>> --- a/meta/classes/kernel-arch.bbclass >>> +++ b/meta/classes/kernel-arch.bbclass >>> @@ -66,3 +66,10 @@ KERNEL_LD = "${CCACHE}${HOST_PREFIX}ld.bfd ${HOST_LD_KERNEL_ARCH}" >>> KERNEL_AR = "${CCACHE}${HOST_PREFIX}ar ${HOST_AR_KERNEL_ARCH}" >>> TOOLCHAIN ?= "gcc" >>> >>> +# 6.3+ requires the variable LOCALVERSION to be set to not get a "+" in >>> +# the local version. Having it empty means nothing will be added, and any >>> +# value will be appended to the local kernel version. This replaces the >>> +# use of .scmversion file for setting a localversion without using >>> +# the CONFIG_LOCALVERSION option. >>> +KERNEL_LOCALVERSION ??= "" >>> +export LOCALVERSION ?= "${KERNEL_LOCALVERSION}" >>> diff --git a/meta/classes/kernel.bbclass b/meta/classes/kernel.bbclass >>> index 5951347361..940f1a3cf4 100644 >>> --- a/meta/classes/kernel.bbclass >>> +++ b/meta/classes/kernel.bbclass >>> @@ -418,7 +418,7 @@ do_compile_kernelmodules() { >>> if (grep -q -i -e '^CONFIG_MODULES=y$' ${B}/.config); then >>> oe_runmake -C ${B} ${PARALLEL_MAKE} modules ${KERNEL_EXTRA_ARGS} >>> >>> - # Module.symvers gets updated during the >>> + # Module.symvers gets updated during the >>> # building of the kernel modules. We need to >>> # update this in the shared workdir since some >>> # external kernel modules has a dependency on >>> @@ -635,7 +635,13 @@ kernel_do_configure() { >>> # $ scripts/setlocalversion . => + >>> # $ make kernelversion => 2.6.37 >>> # $ make kernelrelease => 2.6.37+ >>> - touch ${B}/.scmversion ${S}/.scmversion >>> + # See kernel-arch.bbclass for post v6.3 removal of the extra >>> + # + in localversion. .scmversion is no longer used, and the >>> + # variable LOCALVERSION must be used >>> + if [ ! -e ${B}/.scmversion -a ! -e ${S}/.scmversion ]; then >>> + echo ${KERNEL_LOCALVERSION} > ${B}/.scmversion >>> + echo ${KERNEL_LOCALVERSION} > ${S}/.scmversion >>> + fi >>> >>> if [ "${S}" != "${B}" ] && [ -f "${S}/.config" ] && [ ! -f "${B}/.config" ]; then >>> mv "${S}/.config" "${B}/.config" >>> >>> >>> >>> -=-=-=-=-=-=-=-=-=-=-=- >>> Links: You receive all messages sent to this group. >>> View/Reply Online (#195328): https://lists.openembedded.org/g/openembedded-core/message/195328 >>> Mute This Topic: https://lists.openembedded.org/mt/104311283/6551054 >>> Group Owner: openembedded-core+owner@lists.openembedded.org >>> Unsubscribe: https://lists.openembedded.org/g/openembedded-core/unsub [reatmon@ti.com] >>> -=-=-=-=-=-=-=-=-=-=-=- >>> >> >> -- >> Ryan Eatmon reatmon@ti.com >> ----------------------------------------- >> Texas Instruments, Inc. - LCPD - MGTS -- Ryan Eatmon reatmon@ti.com ----------------------------------------- Texas Instruments, Inc. - LCPD - MGTS ^ permalink raw reply [flat|nested] 33+ messages in thread
* Re: [OE-core][kirkstone 7/8] kernel: fix localversion in v6.3+ 2024-02-16 1:45 ` Ryan Eatmon @ 2024-02-16 8:11 ` Andreas Helbech Kleist 2024-02-16 13:51 ` Andreas Helbech Kleist 2024-02-21 14:39 ` Ryan Eatmon 0 siblings, 2 replies; 33+ messages in thread From: Andreas Helbech Kleist @ 2024-02-16 8:11 UTC (permalink / raw) To: reatmon, Steve Sakoman; +Cc: openembedded-core On Thu, 2024-02-15 at 19:45 -0600, Ryan Eatmon via lists.openembedded.org wrote: > > On 2/15/2024 7:43 PM, Steve Sakoman wrote: > > On Thu, Feb 15, 2024 at 3:20 PM Ryan Eatmon <reatmon@ti.com> wrote: > > > > > > > > > With this patch in place we are seeing a breakage on our kernel builds. > > > This patch was cherry picked from master, but the next patch on this > > > file in master removes the LOCALVERSION setting... > > > > > > https://git.openembedded.org/openembedded-core/commit/meta/classes-recipe/kernel-arch.bbclass?h=master-next&id=b378eec156998eea55ba61e59103cb34fab0d07c > > > > > > > > > There is a disconnect here. Why did we change kirkstone, a stable LTS > > > version with a partial patch series based on master? > > > > The original kirkstone backport request for this patch was sent to the > > list for review on February 9. > > > > There were no comments or objections, so I added it to the patch test > > queue. No issues were encountered on the autobuilder, so I sent this > > patch (along with the rest of the patch queue) to the list for a > > second review opportunity on February 12. Once again there were no > > comments or objections, so it was merged today February 15. > > > > I follow this process because I'm not smart enough, nor do I have time > > enough to thoroughly research all implications of every patch. I rely > > on autobuilder testing and community review to minimize breakage. > > > > This normally works quite well, but this time it didn't. I'll revert > > this patch and if someone would like to resubmit a proper series to > > deal with the issue it is trying to fix I will consider it and run it > > through the same process. > > I can appreciate that. My comment was more for the person who sent the > patch in the first place. You are doing a great job. I will try and > pay attention for a replacement patch coming in and review as well. I'm sorry about that. I need to use a 6.6 kernel on kirkstone, and this patch fixed my usecase. It looked like a simple bugfix, and I was not aware that there was a later fix for this change. Is there any way I can reproduce the error you saw Ryan? And would it make sense to add an automated test for this usecase? I'll test my usecase with both patches backported and re-submit. /Andreas PS: This is my first submission to oe-core, so I'm just starting to learn. I have no idea what or how the autobuilder tests, but I'm willing to learn. > > > > Steve > > > > > > > On 2/12/2024 7:54 AM, Steve Sakoman wrote: > > > > From: Bruce Ashfield <bruce.ashfield@gmail.com> > > > > > > > > During testing of the v6.4 reference kernel, it was noticed that > > > > on-target modules no longer matched the magic value of the running > > > > kernel. > > > > > > > > This was due to a different localversion in the cross built kernel > > > > and the scripts / resources created on target. > > > > > > > > This was due to changes in the setlocalversion script introduced > > > > in the v6.3 series. > > > > > > > > The .scmversion file is no longer used (or packaged) to inhibit > > > > the addition of a "+" (through querying of the git status of the > > > > kernel) or the setting of a local version. > > > > > > > > We recently introduced the KERNEL_LOCALVERSION variable to allow > > > > recipes to place a value in .scmversion, so we extend the use of > > > > that variable to kernel-arch.bbclass and use it to set the > > > > exported variable LOCALVERSION. > > > > > > > > We must do it at the kernel-arch level, as the variable must be > > > > exported in any kernel build to ensure that setlocalversion always > > > > correctly sets the localversion. > > > > > > > > Signed-off-by: Bruce Ashfield <bruce.ashfield@gmail.com> > > > > Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org> > > > > > > > > Cherry-picked from master 765b13b7305c8d2f222cfc66d77c02e6a088c691 > > > > > > > > Signed-off-by: Andreas Helbech Kleist <andreaskleist@gmail.com> > > > > Signed-off-by: Steve Sakoman <steve@sakoman.com> > > > > --- > > > > meta/classes/kernel-arch.bbclass | 7 +++++++ > > > > meta/classes/kernel.bbclass | 10 ++++++++-- > > > > 2 files changed, 15 insertions(+), 2 deletions(-) > > > > > > > > diff --git a/meta/classes/kernel-arch.bbclass b/meta/classes/kernel-arch.bbclass > > > > index 4cd08b96fb..0a79dea0af 100644 > > > > --- a/meta/classes/kernel-arch.bbclass > > > > +++ b/meta/classes/kernel-arch.bbclass > > > > @@ -66,3 +66,10 @@ KERNEL_LD = "${CCACHE}${HOST_PREFIX}ld.bfd ${HOST_LD_KERNEL_ARCH}" > > > > KERNEL_AR = "${CCACHE}${HOST_PREFIX}ar ${HOST_AR_KERNEL_ARCH}" > > > > TOOLCHAIN ?= "gcc" > > > > > > > > +# 6.3+ requires the variable LOCALVERSION to be set to not get a "+" in > > > > +# the local version. Having it empty means nothing will be added, and any > > > > +# value will be appended to the local kernel version. This replaces the > > > > +# use of .scmversion file for setting a localversion without using > > > > +# the CONFIG_LOCALVERSION option. > > > > +KERNEL_LOCALVERSION ??= "" > > > > +export LOCALVERSION ?= "${KERNEL_LOCALVERSION}" > > > > diff --git a/meta/classes/kernel.bbclass b/meta/classes/kernel.bbclass > > > > index 5951347361..940f1a3cf4 100644 > > > > --- a/meta/classes/kernel.bbclass > > > > +++ b/meta/classes/kernel.bbclass > > > > @@ -418,7 +418,7 @@ do_compile_kernelmodules() { > > > > if (grep -q -i -e '^CONFIG_MODULES=y$' ${B}/.config); then > > > > oe_runmake -C ${B} ${PARALLEL_MAKE} modules ${KERNEL_EXTRA_ARGS} > > > > > > > > - # Module.symvers gets updated during the > > > > + # Module.symvers gets updated during the > > > > # building of the kernel modules. We need to > > > > # update this in the shared workdir since some > > > > # external kernel modules has a dependency on > > > > @@ -635,7 +635,13 @@ kernel_do_configure() { > > > > # $ scripts/setlocalversion . => + > > > > # $ make kernelversion => 2.6.37 > > > > # $ make kernelrelease => 2.6.37+ > > > > - touch ${B}/.scmversion ${S}/.scmversion > > > > + # See kernel-arch.bbclass for post v6.3 removal of the extra > > > > + # + in localversion. .scmversion is no longer used, and the > > > > + # variable LOCALVERSION must be used > > > > + if [ ! -e ${B}/.scmversion -a ! -e ${S}/.scmversion ]; then > > > > + echo ${KERNEL_LOCALVERSION} > ${B}/.scmversion > > > > + echo ${KERNEL_LOCALVERSION} > ${S}/.scmversion > > > > + fi > > > > > > > > if [ "${S}" != "${B}" ] && [ -f "${S}/.config" ] && [ ! -f "${B}/.config" ]; then > > > > mv "${S}/.config" "${B}/.config" > > > > > > > > > > > > > > > > > > > > > > > > > > -- > > > Ryan Eatmon reatmon@ti.com > > > ----------------------------------------- > > > Texas Instruments, Inc. - LCPD - MGTS > > -=-=-=-=-=-=-=-=-=-=-=- > Links: You receive all messages sent to this group. > View/Reply Online (#195736): https://lists.openembedded.org/g/openembedded-core/message/195736 > Mute This Topic: https://lists.openembedded.org/mt/104311283/7501392 > Group Owner: openembedded-core+owner@lists.openembedded.org > Unsubscribe: https://lists.openembedded.org/g/openembedded-core/unsub [andreaskleist@gmail.com] > -=-=-=-=-=-=-=-=-=-=-=- > ^ permalink raw reply [flat|nested] 33+ messages in thread
* Re: [OE-core][kirkstone 7/8] kernel: fix localversion in v6.3+ 2024-02-16 8:11 ` Andreas Helbech Kleist @ 2024-02-16 13:51 ` Andreas Helbech Kleist 2024-02-21 14:40 ` Ryan Eatmon 2024-02-21 14:39 ` Ryan Eatmon 1 sibling, 1 reply; 33+ messages in thread From: Andreas Helbech Kleist @ 2024-02-16 13:51 UTC (permalink / raw) To: reatmon, Steve Sakoman; +Cc: openembedded-core On Fri, 2024-02-16 at 09:11 +0100, Andreas Helbech Kleist wrote: > On Thu, 2024-02-15 at 19:45 -0600, Ryan Eatmon via > lists.openembedded.org wrote: > > > > On 2/15/2024 7:43 PM, Steve Sakoman wrote: > > > On Thu, Feb 15, 2024 at 3:20 PM Ryan Eatmon <reatmon@ti.com> wrote: > > > > > > > > > > > > With this patch in place we are seeing a breakage on our kernel builds. > > > > This patch was cherry picked from master, but the next patch on this > > > > file in master removes the LOCALVERSION setting... > > > > > > > > https://git.openembedded.org/openembedded-core/commit/meta/classes-recipe/kernel-arch.bbclass?h=master-next&id=b378eec156998eea55ba61e59103cb34fab0d07c > > > > > > > > > > > > There is a disconnect here. Why did we change kirkstone, a stable LTS > > > > version with a partial patch series based on master? > > > > > > The original kirkstone backport request for this patch was sent to the > > > list for review on February 9. > > > > > > There were no comments or objections, so I added it to the patch test > > > queue. No issues were encountered on the autobuilder, so I sent this > > > patch (along with the rest of the patch queue) to the list for a > > > second review opportunity on February 12. Once again there were no > > > comments or objections, so it was merged today February 15. > > > > > > I follow this process because I'm not smart enough, nor do I have time > > > enough to thoroughly research all implications of every patch. I rely > > > on autobuilder testing and community review to minimize breakage. > > > > > > This normally works quite well, but this time it didn't. I'll revert > > > this patch and if someone would like to resubmit a proper series to > > > deal with the issue it is trying to fix I will consider it and run it > > > through the same process. > > > > I can appreciate that. My comment was more for the person who sent the > > patch in the first place. You are doing a great job. I will try and > > pay attention for a replacement patch coming in and review as well. > > I'm sorry about that. I need to use a 6.6 kernel on kirkstone, and this > patch fixed my usecase. It looked like a simple bugfix, and I was not > aware that there was a later fix for this change. > > Is there any way I can reproduce the error you saw Ryan? > > And would it make sense to add an automated test for this usecase? > > I'll test my usecase with both patches backported and re-submit. I've now looked into it and can make it work in my setup by backporting these three commits from master: kernel.bbclass: introduce KERNEL_LOCALVERSION https://git.openembedded.org/openembedded-core/commit/?h=master-next&id=229435a52f36ddec5f85fb6d5ccd42044b688397 kernel: fix localversion in v6.3+ https://git.openembedded.org/openembedded-core/commit/?h=master-next&id=765b13b7305c8d2f222cfc66d77c02e6a088c691 kernel: make LOCALVERSION consistent between recipes https://git.openembedded.org/openembedded-core/commit/?h=master-next&id=b378eec156998eea55ba61e59103cb34fab0d07c The cherry-pick's apply cleanly without any conflicts. Would they be appropriate for kirkstone? I'm wondering if there would be any issue with backporting the KERNEL_LOCALVERSION variable specifically. If I hear nothing before the middle of next week, I'll submit the patch series (CC'ing you Ryan). /Andreas > > PS: This is my first submission to oe-core, so I'm just starting to > learn. I have no idea what or how the autobuilder tests, but I'm > willing to learn. > > > > > > > > Steve > > > > > > > > > > On 2/12/2024 7:54 AM, Steve Sakoman wrote: > > > > > From: Bruce Ashfield <bruce.ashfield@gmail.com> > > > > > > > > > > During testing of the v6.4 reference kernel, it was noticed that > > > > > on-target modules no longer matched the magic value of the running > > > > > kernel. > > > > > > > > > > This was due to a different localversion in the cross built kernel > > > > > and the scripts / resources created on target. > > > > > > > > > > This was due to changes in the setlocalversion script introduced > > > > > in the v6.3 series. > > > > > > > > > > The .scmversion file is no longer used (or packaged) to inhibit > > > > > the addition of a "+" (through querying of the git status of the > > > > > kernel) or the setting of a local version. > > > > > > > > > > We recently introduced the KERNEL_LOCALVERSION variable to allow > > > > > recipes to place a value in .scmversion, so we extend the use of > > > > > that variable to kernel-arch.bbclass and use it to set the > > > > > exported variable LOCALVERSION. > > > > > > > > > > We must do it at the kernel-arch level, as the variable must be > > > > > exported in any kernel build to ensure that setlocalversion always > > > > > correctly sets the localversion. > > > > > > > > > > Signed-off-by: Bruce Ashfield <bruce.ashfield@gmail.com> > > > > > Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org> > > > > > > > > > > Cherry-picked from master 765b13b7305c8d2f222cfc66d77c02e6a088c691 > > > > > > > > > > Signed-off-by: Andreas Helbech Kleist <andreaskleist@gmail.com> > > > > > Signed-off-by: Steve Sakoman <steve@sakoman.com> > > > > > --- > > > > > meta/classes/kernel-arch.bbclass | 7 +++++++ > > > > > meta/classes/kernel.bbclass | 10 ++++++++-- > > > > > 2 files changed, 15 insertions(+), 2 deletions(-) > > > > > > > > > > diff --git a/meta/classes/kernel-arch.bbclass b/meta/classes/kernel-arch.bbclass > > > > > index 4cd08b96fb..0a79dea0af 100644 > > > > > --- a/meta/classes/kernel-arch.bbclass > > > > > +++ b/meta/classes/kernel-arch.bbclass > > > > > @@ -66,3 +66,10 @@ KERNEL_LD = "${CCACHE}${HOST_PREFIX}ld.bfd ${HOST_LD_KERNEL_ARCH}" > > > > > KERNEL_AR = "${CCACHE}${HOST_PREFIX}ar ${HOST_AR_KERNEL_ARCH}" > > > > > TOOLCHAIN ?= "gcc" > > > > > > > > > > +# 6.3+ requires the variable LOCALVERSION to be set to not get a "+" in > > > > > +# the local version. Having it empty means nothing will be added, and any > > > > > +# value will be appended to the local kernel version. This replaces the > > > > > +# use of .scmversion file for setting a localversion without using > > > > > +# the CONFIG_LOCALVERSION option. > > > > > +KERNEL_LOCALVERSION ??= "" > > > > > +export LOCALVERSION ?= "${KERNEL_LOCALVERSION}" > > > > > diff --git a/meta/classes/kernel.bbclass b/meta/classes/kernel.bbclass > > > > > index 5951347361..940f1a3cf4 100644 > > > > > --- a/meta/classes/kernel.bbclass > > > > > +++ b/meta/classes/kernel.bbclass > > > > > @@ -418,7 +418,7 @@ do_compile_kernelmodules() { > > > > > if (grep -q -i -e '^CONFIG_MODULES=y$' ${B}/.config); then > > > > > oe_runmake -C ${B} ${PARALLEL_MAKE} modules ${KERNEL_EXTRA_ARGS} > > > > > > > > > > - # Module.symvers gets updated during the > > > > > + # Module.symvers gets updated during the > > > > > # building of the kernel modules. We need to > > > > > # update this in the shared workdir since some > > > > > # external kernel modules has a dependency on > > > > > @@ -635,7 +635,13 @@ kernel_do_configure() { > > > > > # $ scripts/setlocalversion . => + > > > > > # $ make kernelversion => 2.6.37 > > > > > # $ make kernelrelease => 2.6.37+ > > > > > - touch ${B}/.scmversion ${S}/.scmversion > > > > > + # See kernel-arch.bbclass for post v6.3 removal of the extra > > > > > + # + in localversion. .scmversion is no longer used, and the > > > > > + # variable LOCALVERSION must be used > > > > > + if [ ! -e ${B}/.scmversion -a ! -e ${S}/.scmversion ]; then > > > > > + echo ${KERNEL_LOCALVERSION} > ${B}/.scmversion > > > > > + echo ${KERNEL_LOCALVERSION} > ${S}/.scmversion > > > > > + fi > > > > > > > > > > if [ "${S}" != "${B}" ] && [ -f "${S}/.config" ] && [ ! -f "${B}/.config" ]; then > > > > > mv "${S}/.config" "${B}/.config" > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > -- > > > > Ryan Eatmon reatmon@ti.com > > > > ----------------------------------------- > > > > Texas Instruments, Inc. - LCPD - MGTS > > > > -=-=-=-=-=-=-=-=-=-=-=- > > Links: You receive all messages sent to this group. > > View/Reply Online (#195736): https://lists.openembedded.org/g/openembedded-core/message/195736 > > Mute This Topic: https://lists.openembedded.org/mt/104311283/7501392 > > Group Owner: openembedded-core+owner@lists.openembedded.org > > Unsubscribe: https://lists.openembedded.org/g/openembedded-core/unsub [andreaskleist@gmail.com] > > -=-=-=-=-=-=-=-=-=-=-=- > > > ^ permalink raw reply [flat|nested] 33+ messages in thread
* Re: [OE-core][kirkstone 7/8] kernel: fix localversion in v6.3+ 2024-02-16 13:51 ` Andreas Helbech Kleist @ 2024-02-21 14:40 ` Ryan Eatmon 2024-02-22 10:39 ` Andreas Helbech Kleist 0 siblings, 1 reply; 33+ messages in thread From: Ryan Eatmon @ 2024-02-21 14:40 UTC (permalink / raw) To: Andreas Helbech Kleist, Steve Sakoman; +Cc: openembedded-core On 2/16/2024 7:51 AM, Andreas Helbech Kleist wrote: > On Fri, 2024-02-16 at 09:11 +0100, Andreas Helbech Kleist wrote: >> On Thu, 2024-02-15 at 19:45 -0600, Ryan Eatmon via >> lists.openembedded.org wrote: >>> >>> On 2/15/2024 7:43 PM, Steve Sakoman wrote: >>>> On Thu, Feb 15, 2024 at 3:20 PM Ryan Eatmon <reatmon@ti.com> wrote: >>>>> >>>>> >>>>> With this patch in place we are seeing a breakage on our kernel builds. >>>>> This patch was cherry picked from master, but the next patch on this >>>>> file in master removes the LOCALVERSION setting... >>>>> >>>>> https://git.openembedded.org/openembedded-core/commit/meta/classes-recipe/kernel-arch.bbclass?h=master-next&id=b378eec156998eea55ba61e59103cb34fab0d07c >>>>> >>>>> >>>>> There is a disconnect here. Why did we change kirkstone, a stable LTS >>>>> version with a partial patch series based on master? >>>> >>>> The original kirkstone backport request for this patch was sent to the >>>> list for review on February 9. >>>> >>>> There were no comments or objections, so I added it to the patch test >>>> queue. No issues were encountered on the autobuilder, so I sent this >>>> patch (along with the rest of the patch queue) to the list for a >>>> second review opportunity on February 12. Once again there were no >>>> comments or objections, so it was merged today February 15. >>>> >>>> I follow this process because I'm not smart enough, nor do I have time >>>> enough to thoroughly research all implications of every patch. I rely >>>> on autobuilder testing and community review to minimize breakage. >>>> >>>> This normally works quite well, but this time it didn't. I'll revert >>>> this patch and if someone would like to resubmit a proper series to >>>> deal with the issue it is trying to fix I will consider it and run it >>>> through the same process. >>> >>> I can appreciate that. My comment was more for the person who sent the >>> patch in the first place. You are doing a great job. I will try and >>> pay attention for a replacement patch coming in and review as well. >> >> I'm sorry about that. I need to use a 6.6 kernel on kirkstone, and this >> patch fixed my usecase. It looked like a simple bugfix, and I was not >> aware that there was a later fix for this change. >> >> Is there any way I can reproduce the error you saw Ryan? >> >> And would it make sense to add an automated test for this usecase? >> >> I'll test my usecase with both patches backported and re-submit. > > I've now looked into it and can make it work in my setup by backporting > these three commits from master: > > kernel.bbclass: introduce KERNEL_LOCALVERSION > > https://git.openembedded.org/openembedded-core/commit/?h=master-next&id=229435a52f36ddec5f85fb6d5ccd42044b688397 > > kernel: fix localversion in v6.3+ > > https://git.openembedded.org/openembedded-core/commit/?h=master-next&id=765b13b7305c8d2f222cfc66d77c02e6a088c691 > > kernel: make LOCALVERSION consistent between recipes > > https://git.openembedded.org/openembedded-core/commit/?h=master-next&id=b378eec156998eea55ba61e59103cb34fab0d07c > > The cherry-pick's apply cleanly without any conflicts. > > Would they be appropriate for kirkstone? I'm wondering if there would > be any issue with backporting the KERNEL_LOCALVERSION variable > specifically. > > If I hear nothing before the middle of next week, I'll submit the patch > series (CC'ing you Ryan). > Can you send me the patch series that you want to want to send in against kirkstone? I'll apply them locally and run some tests on our setup to make sure they don't cause an issue. I don't foresee them being a problem, but I'm more than happy to verify. > /Andreas >> >> PS: This is my first submission to oe-core, so I'm just starting to >> learn. I have no idea what or how the autobuilder tests, but I'm >> willing to learn. >> >>> >>> >>>> Steve >>>> >>>> >>>>> On 2/12/2024 7:54 AM, Steve Sakoman wrote: >>>>>> From: Bruce Ashfield <bruce.ashfield@gmail.com> >>>>>> >>>>>> During testing of the v6.4 reference kernel, it was noticed that >>>>>> on-target modules no longer matched the magic value of the running >>>>>> kernel. >>>>>> >>>>>> This was due to a different localversion in the cross built kernel >>>>>> and the scripts / resources created on target. >>>>>> >>>>>> This was due to changes in the setlocalversion script introduced >>>>>> in the v6.3 series. >>>>>> >>>>>> The .scmversion file is no longer used (or packaged) to inhibit >>>>>> the addition of a "+" (through querying of the git status of the >>>>>> kernel) or the setting of a local version. >>>>>> >>>>>> We recently introduced the KERNEL_LOCALVERSION variable to allow >>>>>> recipes to place a value in .scmversion, so we extend the use of >>>>>> that variable to kernel-arch.bbclass and use it to set the >>>>>> exported variable LOCALVERSION. >>>>>> >>>>>> We must do it at the kernel-arch level, as the variable must be >>>>>> exported in any kernel build to ensure that setlocalversion always >>>>>> correctly sets the localversion. >>>>>> >>>>>> Signed-off-by: Bruce Ashfield <bruce.ashfield@gmail.com> >>>>>> Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org> >>>>>> >>>>>> Cherry-picked from master 765b13b7305c8d2f222cfc66d77c02e6a088c691 >>>>>> >>>>>> Signed-off-by: Andreas Helbech Kleist <andreaskleist@gmail.com> >>>>>> Signed-off-by: Steve Sakoman <steve@sakoman.com> >>>>>> --- >>>>>> meta/classes/kernel-arch.bbclass | 7 +++++++ >>>>>> meta/classes/kernel.bbclass | 10 ++++++++-- >>>>>> 2 files changed, 15 insertions(+), 2 deletions(-) >>>>>> >>>>>> diff --git a/meta/classes/kernel-arch.bbclass b/meta/classes/kernel-arch.bbclass >>>>>> index 4cd08b96fb..0a79dea0af 100644 >>>>>> --- a/meta/classes/kernel-arch.bbclass >>>>>> +++ b/meta/classes/kernel-arch.bbclass >>>>>> @@ -66,3 +66,10 @@ KERNEL_LD = "${CCACHE}${HOST_PREFIX}ld.bfd ${HOST_LD_KERNEL_ARCH}" >>>>>> KERNEL_AR = "${CCACHE}${HOST_PREFIX}ar ${HOST_AR_KERNEL_ARCH}" >>>>>> TOOLCHAIN ?= "gcc" >>>>>> >>>>>> +# 6.3+ requires the variable LOCALVERSION to be set to not get a "+" in >>>>>> +# the local version. Having it empty means nothing will be added, and any >>>>>> +# value will be appended to the local kernel version. This replaces the >>>>>> +# use of .scmversion file for setting a localversion without using >>>>>> +# the CONFIG_LOCALVERSION option. >>>>>> +KERNEL_LOCALVERSION ??= "" >>>>>> +export LOCALVERSION ?= "${KERNEL_LOCALVERSION}" >>>>>> diff --git a/meta/classes/kernel.bbclass b/meta/classes/kernel.bbclass >>>>>> index 5951347361..940f1a3cf4 100644 >>>>>> --- a/meta/classes/kernel.bbclass >>>>>> +++ b/meta/classes/kernel.bbclass >>>>>> @@ -418,7 +418,7 @@ do_compile_kernelmodules() { >>>>>> if (grep -q -i -e '^CONFIG_MODULES=y$' ${B}/.config); then >>>>>> oe_runmake -C ${B} ${PARALLEL_MAKE} modules ${KERNEL_EXTRA_ARGS} >>>>>> >>>>>> - # Module.symvers gets updated during the >>>>>> + # Module.symvers gets updated during the >>>>>> # building of the kernel modules. We need to >>>>>> # update this in the shared workdir since some >>>>>> # external kernel modules has a dependency on >>>>>> @@ -635,7 +635,13 @@ kernel_do_configure() { >>>>>> # $ scripts/setlocalversion . => + >>>>>> # $ make kernelversion => 2.6.37 >>>>>> # $ make kernelrelease => 2.6.37+ >>>>>> - touch ${B}/.scmversion ${S}/.scmversion >>>>>> + # See kernel-arch.bbclass for post v6.3 removal of the extra >>>>>> + # + in localversion. .scmversion is no longer used, and the >>>>>> + # variable LOCALVERSION must be used >>>>>> + if [ ! -e ${B}/.scmversion -a ! -e ${S}/.scmversion ]; then >>>>>> + echo ${KERNEL_LOCALVERSION} > ${B}/.scmversion >>>>>> + echo ${KERNEL_LOCALVERSION} > ${S}/.scmversion >>>>>> + fi >>>>>> >>>>>> if [ "${S}" != "${B}" ] && [ -f "${S}/.config" ] && [ ! -f "${B}/.config" ]; then >>>>>> mv "${S}/.config" "${B}/.config" >>>>>> >>>>>> >>>>>> >>>>>> >>>>>> >>>>> >>>>> -- >>>>> Ryan Eatmon reatmon@ti.com >>>>> ----------------------------------------- >>>>> Texas Instruments, Inc. - LCPD - MGTS >>> >>> -=-=-=-=-=-=-=-=-=-=-=- >>> Links: You receive all messages sent to this group. >>> View/Reply Online (#195736): https://lists.openembedded.org/g/openembedded-core/message/195736 >>> Mute This Topic: https://lists.openembedded.org/mt/104311283/7501392 >>> Group Owner: openembedded-core+owner@lists.openembedded.org >>> Unsubscribe: https://lists.openembedded.org/g/openembedded-core/unsub [andreaskleist@gmail.com] >>> -=-=-=-=-=-=-=-=-=-=-=- >>> >> > -- Ryan Eatmon reatmon@ti.com ----------------------------------------- Texas Instruments, Inc. - LCPD - MGTS ^ permalink raw reply [flat|nested] 33+ messages in thread
* Re: [OE-core][kirkstone 7/8] kernel: fix localversion in v6.3+ 2024-02-21 14:40 ` Ryan Eatmon @ 2024-02-22 10:39 ` Andreas Helbech Kleist 2024-02-22 18:38 ` Ryan Eatmon 0 siblings, 1 reply; 33+ messages in thread From: Andreas Helbech Kleist @ 2024-02-22 10:39 UTC (permalink / raw) To: Ryan Eatmon, Steve Sakoman; +Cc: openembedded-core On Wed, 2024-02-21 at 08:40 -0600, Ryan Eatmon wrote: > > On 2/16/2024 7:51 AM, Andreas Helbech Kleist wrote: > > On Fri, 2024-02-16 at 09:11 +0100, Andreas Helbech Kleist wrote: > > > On Thu, 2024-02-15 at 19:45 -0600, Ryan Eatmon via > > > lists.openembedded.org wrote: > > > > > > > > On 2/15/2024 7:43 PM, Steve Sakoman wrote: > > > > > On Thu, Feb 15, 2024 at 3:20 PM Ryan Eatmon <reatmon@ti.com> wrote: > > > > > > > > > > > > > > > > > > With this patch in place we are seeing a breakage on our kernel builds. > > > > > > This patch was cherry picked from master, but the next patch on this > > > > > > file in master removes the LOCALVERSION setting... > > > > > > > > > > > > https://git.openembedded.org/openembedded-core/commit/meta/classes-recipe/kernel-arch.bbclass?h=master-next&id=b378eec156998eea55ba61e59103cb34fab0d07c > > > > > > > > > > > > > > > > > > There is a disconnect here. Why did we change kirkstone, a stable LTS > > > > > > version with a partial patch series based on master? > > > > > > > > > > The original kirkstone backport request for this patch was sent to the > > > > > list for review on February 9. > > > > > > > > > > There were no comments or objections, so I added it to the patch test > > > > > queue. No issues were encountered on the autobuilder, so I sent this > > > > > patch (along with the rest of the patch queue) to the list for a > > > > > second review opportunity on February 12. Once again there were no > > > > > comments or objections, so it was merged today February 15. > > > > > > > > > > I follow this process because I'm not smart enough, nor do I have time > > > > > enough to thoroughly research all implications of every patch. I rely > > > > > on autobuilder testing and community review to minimize breakage. > > > > > > > > > > This normally works quite well, but this time it didn't. I'll revert > > > > > this patch and if someone would like to resubmit a proper series to > > > > > deal with the issue it is trying to fix I will consider it and run it > > > > > through the same process. > > > > > > > > I can appreciate that. My comment was more for the person who sent the > > > > patch in the first place. You are doing a great job. I will try and > > > > pay attention for a replacement patch coming in and review as well. > > > > > > I'm sorry about that. I need to use a 6.6 kernel on kirkstone, and this > > > patch fixed my usecase. It looked like a simple bugfix, and I was not > > > aware that there was a later fix for this change. > > > > > > Is there any way I can reproduce the error you saw Ryan? > > > > > > And would it make sense to add an automated test for this usecase? > > > > > > I'll test my usecase with both patches backported and re-submit. > > > > I've now looked into it and can make it work in my setup by backporting > > these three commits from master: > > > > kernel.bbclass: introduce KERNEL_LOCALVERSION > > > > https://git.openembedded.org/openembedded-core/commit/?h=master-next&id=229435a52f36ddec5f85fb6d5ccd42044b688397 > > > > kernel: fix localversion in v6.3+ > > > > https://git.openembedded.org/openembedded-core/commit/?h=master-next&id=765b13b7305c8d2f222cfc66d77c02e6a088c691 > > > > kernel: make LOCALVERSION consistent between recipes > > > > https://git.openembedded.org/openembedded-core/commit/?h=master-next&id=b378eec156998eea55ba61e59103cb34fab0d07c > > > > The cherry-pick's apply cleanly without any conflicts. > > > > Would they be appropriate for kirkstone? I'm wondering if there would > > be any issue with backporting the KERNEL_LOCALVERSION variable > > specifically. > > > > If I hear nothing before the middle of next week, I'll submit the patch > > series (CC'ing you Ryan). > > > > Can you send me the patch series that you want to want to send in > against kirkstone? I'll apply them locally and run some tests on our > setup to make sure they don't cause an issue. I don't foresee them > being a problem, but I'm more than happy to verify. I send them to the list and CC'ed you, hope that is ok. See https://lore.kernel.org/openembedded-core/20240222103545.2970387-1-andreaskleist@gmail.com/T/#t > > /Andreas > > > > > > PS: This is my first submission to oe-core, so I'm just starting to > > > learn. I have no idea what or how the autobuilder tests, but I'm > > > willing to learn. > > > > > > > > > > > > > > > > Steve > > > > > > > > > > > > > > > > On 2/12/2024 7:54 AM, Steve Sakoman wrote: > > > > > > > From: Bruce Ashfield <bruce.ashfield@gmail.com> > > > > > > > > > > > > > > During testing of the v6.4 reference kernel, it was noticed that > > > > > > > on-target modules no longer matched the magic value of the running > > > > > > > kernel. > > > > > > > > > > > > > > This was due to a different localversion in the cross built kernel > > > > > > > and the scripts / resources created on target. > > > > > > > > > > > > > > This was due to changes in the setlocalversion script introduced > > > > > > > in the v6.3 series. > > > > > > > > > > > > > > The .scmversion file is no longer used (or packaged) to inhibit > > > > > > > the addition of a "+" (through querying of the git status of the > > > > > > > kernel) or the setting of a local version. > > > > > > > > > > > > > > We recently introduced the KERNEL_LOCALVERSION variable to allow > > > > > > > recipes to place a value in .scmversion, so we extend the use of > > > > > > > that variable to kernel-arch.bbclass and use it to set the > > > > > > > exported variable LOCALVERSION. > > > > > > > > > > > > > > We must do it at the kernel-arch level, as the variable must be > > > > > > > exported in any kernel build to ensure that setlocalversion always > > > > > > > correctly sets the localversion. > > > > > > > > > > > > > > Signed-off-by: Bruce Ashfield <bruce.ashfield@gmail.com> > > > > > > > Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org> > > > > > > > > > > > > > > Cherry-picked from master 765b13b7305c8d2f222cfc66d77c02e6a088c691 > > > > > > > > > > > > > > Signed-off-by: Andreas Helbech Kleist <andreaskleist@gmail.com> > > > > > > > Signed-off-by: Steve Sakoman <steve@sakoman.com> > > > > > > > --- > > > > > > > meta/classes/kernel-arch.bbclass | 7 +++++++ > > > > > > > meta/classes/kernel.bbclass | 10 ++++++++-- > > > > > > > 2 files changed, 15 insertions(+), 2 deletions(-) > > > > > > > > > > > > > > diff --git a/meta/classes/kernel-arch.bbclass b/meta/classes/kernel-arch.bbclass > > > > > > > index 4cd08b96fb..0a79dea0af 100644 > > > > > > > --- a/meta/classes/kernel-arch.bbclass > > > > > > > +++ b/meta/classes/kernel-arch.bbclass > > > > > > > @@ -66,3 +66,10 @@ KERNEL_LD = "${CCACHE}${HOST_PREFIX}ld.bfd ${HOST_LD_KERNEL_ARCH}" > > > > > > > KERNEL_AR = "${CCACHE}${HOST_PREFIX}ar ${HOST_AR_KERNEL_ARCH}" > > > > > > > TOOLCHAIN ?= "gcc" > > > > > > > > > > > > > > +# 6.3+ requires the variable LOCALVERSION to be set to not get a "+" in > > > > > > > +# the local version. Having it empty means nothing will be added, and any > > > > > > > +# value will be appended to the local kernel version. This replaces the > > > > > > > +# use of .scmversion file for setting a localversion without using > > > > > > > +# the CONFIG_LOCALVERSION option. > > > > > > > +KERNEL_LOCALVERSION ??= "" > > > > > > > +export LOCALVERSION ?= "${KERNEL_LOCALVERSION}" > > > > > > > diff --git a/meta/classes/kernel.bbclass b/meta/classes/kernel.bbclass > > > > > > > index 5951347361..940f1a3cf4 100644 > > > > > > > --- a/meta/classes/kernel.bbclass > > > > > > > +++ b/meta/classes/kernel.bbclass > > > > > > > @@ -418,7 +418,7 @@ do_compile_kernelmodules() { > > > > > > > if (grep -q -i -e '^CONFIG_MODULES=y$' ${B}/.config); then > > > > > > > oe_runmake -C ${B} ${PARALLEL_MAKE} modules ${KERNEL_EXTRA_ARGS} > > > > > > > > > > > > > > - # Module.symvers gets updated during the > > > > > > > + # Module.symvers gets updated during the > > > > > > > # building of the kernel modules. We need to > > > > > > > # update this in the shared workdir since some > > > > > > > # external kernel modules has a dependency on > > > > > > > @@ -635,7 +635,13 @@ kernel_do_configure() { > > > > > > > # $ scripts/setlocalversion . => + > > > > > > > # $ make kernelversion => 2.6.37 > > > > > > > # $ make kernelrelease => 2.6.37+ > > > > > > > - touch ${B}/.scmversion ${S}/.scmversion > > > > > > > + # See kernel-arch.bbclass for post v6.3 removal of the extra > > > > > > > + # + in localversion. .scmversion is no longer used, and the > > > > > > > + # variable LOCALVERSION must be used > > > > > > > + if [ ! -e ${B}/.scmversion -a ! -e ${S}/.scmversion ]; then > > > > > > > + echo ${KERNEL_LOCALVERSION} > ${B}/.scmversion > > > > > > > + echo ${KERNEL_LOCALVERSION} > ${S}/.scmversion > > > > > > > + fi > > > > > > > > > > > > > > if [ "${S}" != "${B}" ] && [ -f "${S}/.config" ] && [ ! -f "${B}/.config" ]; then > > > > > > > mv "${S}/.config" "${B}/.config" > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > -- > > > > > > Ryan Eatmon reatmon@ti.com > > > > > > ----------------------------------------- > > > > > > Texas Instruments, Inc. - LCPD - MGTS > > > > > > > > -=-=-=-=-=-=-=-=-=-=-=- > > > > Links: You receive all messages sent to this group. > > > > View/Reply Online (#195736): https://lists.openembedded.org/g/openembedded-core/message/195736 > > > > Mute This Topic: https://lists.openembedded.org/mt/104311283/7501392 > > > > Group Owner: openembedded-core+owner@lists.openembedded.org > > > > Unsubscribe: https://lists.openembedded.org/g/openembedded-core/unsub [andreaskleist@gmail.com] > > > > -=-=-=-=-=-=-=-=-=-=-=- > > > > > > > > > > ^ permalink raw reply [flat|nested] 33+ messages in thread
* Re: [OE-core][kirkstone 7/8] kernel: fix localversion in v6.3+ 2024-02-22 10:39 ` Andreas Helbech Kleist @ 2024-02-22 18:38 ` Ryan Eatmon 0 siblings, 0 replies; 33+ messages in thread From: Ryan Eatmon @ 2024-02-22 18:38 UTC (permalink / raw) To: Andreas Helbech Kleist, Steve Sakoman; +Cc: openembedded-core On 2/22/2024 4:39 AM, Andreas Helbech Kleist wrote: > On Wed, 2024-02-21 at 08:40 -0600, Ryan Eatmon wrote: >> >> On 2/16/2024 7:51 AM, Andreas Helbech Kleist wrote: >>> On Fri, 2024-02-16 at 09:11 +0100, Andreas Helbech Kleist wrote: >>>> On Thu, 2024-02-15 at 19:45 -0600, Ryan Eatmon via >>>> lists.openembedded.org wrote: >>>>> >>>>> On 2/15/2024 7:43 PM, Steve Sakoman wrote: >>>>>> On Thu, Feb 15, 2024 at 3:20 PM Ryan Eatmon <reatmon@ti.com> wrote: >>>>>>> >>>>>>> >>>>>>> With this patch in place we are seeing a breakage on our kernel builds. >>>>>>> This patch was cherry picked from master, but the next patch on this >>>>>>> file in master removes the LOCALVERSION setting... >>>>>>> >>>>>>> https://git.openembedded.org/openembedded-core/commit/meta/classes-recipe/kernel-arch.bbclass?h=master-next&id=b378eec156998eea55ba61e59103cb34fab0d07c >>>>>>> >>>>>>> >>>>>>> There is a disconnect here. Why did we change kirkstone, a stable LTS >>>>>>> version with a partial patch series based on master? >>>>>> >>>>>> The original kirkstone backport request for this patch was sent to the >>>>>> list for review on February 9. >>>>>> >>>>>> There were no comments or objections, so I added it to the patch test >>>>>> queue. No issues were encountered on the autobuilder, so I sent this >>>>>> patch (along with the rest of the patch queue) to the list for a >>>>>> second review opportunity on February 12. Once again there were no >>>>>> comments or objections, so it was merged today February 15. >>>>>> >>>>>> I follow this process because I'm not smart enough, nor do I have time >>>>>> enough to thoroughly research all implications of every patch. I rely >>>>>> on autobuilder testing and community review to minimize breakage. >>>>>> >>>>>> This normally works quite well, but this time it didn't. I'll revert >>>>>> this patch and if someone would like to resubmit a proper series to >>>>>> deal with the issue it is trying to fix I will consider it and run it >>>>>> through the same process. >>>>> >>>>> I can appreciate that. My comment was more for the person who sent the >>>>> patch in the first place. You are doing a great job. I will try and >>>>> pay attention for a replacement patch coming in and review as well. >>>> >>>> I'm sorry about that. I need to use a 6.6 kernel on kirkstone, and this >>>> patch fixed my usecase. It looked like a simple bugfix, and I was not >>>> aware that there was a later fix for this change. >>>> >>>> Is there any way I can reproduce the error you saw Ryan? >>>> >>>> And would it make sense to add an automated test for this usecase? >>>> >>>> I'll test my usecase with both patches backported and re-submit. >>> >>> I've now looked into it and can make it work in my setup by backporting >>> these three commits from master: >>> >>> kernel.bbclass: introduce KERNEL_LOCALVERSION >>> >>> https://git.openembedded.org/openembedded-core/commit/?h=master-next&id=229435a52f36ddec5f85fb6d5ccd42044b688397 >>> >>> kernel: fix localversion in v6.3+ >>> >>> https://git.openembedded.org/openembedded-core/commit/?h=master-next&id=765b13b7305c8d2f222cfc66d77c02e6a088c691 >>> >>> kernel: make LOCALVERSION consistent between recipes >>> >>> https://git.openembedded.org/openembedded-core/commit/?h=master-next&id=b378eec156998eea55ba61e59103cb34fab0d07c >>> >>> The cherry-pick's apply cleanly without any conflicts. >>> >>> Would they be appropriate for kirkstone? I'm wondering if there would >>> be any issue with backporting the KERNEL_LOCALVERSION variable >>> specifically. >>> >>> If I hear nothing before the middle of next week, I'll submit the patch >>> series (CC'ing you Ryan). >>> >> >> Can you send me the patch series that you want to want to send in >> against kirkstone? I'll apply them locally and run some tests on our >> setup to make sure they don't cause an issue. I don't foresee them >> being a problem, but I'm more than happy to verify. > > I send them to the list and CC'ed you, hope that is ok. Yep. I'll test them and reply to the patches. > See > https://lore.kernel.org/openembedded-core/20240222103545.2970387-1-andreaskleist@gmail.com/T/#t > >>> /Andreas >>>> >>>> PS: This is my first submission to oe-core, so I'm just starting to >>>> learn. I have no idea what or how the autobuilder tests, but I'm >>>> willing to learn. >>>> >>>>> >>>>> >>>>>> Steve >>>>>> >>>>>> >>>>>>> On 2/12/2024 7:54 AM, Steve Sakoman wrote: >>>>>>>> From: Bruce Ashfield <bruce.ashfield@gmail.com> >>>>>>>> >>>>>>>> During testing of the v6.4 reference kernel, it was noticed that >>>>>>>> on-target modules no longer matched the magic value of the running >>>>>>>> kernel. >>>>>>>> >>>>>>>> This was due to a different localversion in the cross built kernel >>>>>>>> and the scripts / resources created on target. >>>>>>>> >>>>>>>> This was due to changes in the setlocalversion script introduced >>>>>>>> in the v6.3 series. >>>>>>>> >>>>>>>> The .scmversion file is no longer used (or packaged) to inhibit >>>>>>>> the addition of a "+" (through querying of the git status of the >>>>>>>> kernel) or the setting of a local version. >>>>>>>> >>>>>>>> We recently introduced the KERNEL_LOCALVERSION variable to allow >>>>>>>> recipes to place a value in .scmversion, so we extend the use of >>>>>>>> that variable to kernel-arch.bbclass and use it to set the >>>>>>>> exported variable LOCALVERSION. >>>>>>>> >>>>>>>> We must do it at the kernel-arch level, as the variable must be >>>>>>>> exported in any kernel build to ensure that setlocalversion always >>>>>>>> correctly sets the localversion. >>>>>>>> >>>>>>>> Signed-off-by: Bruce Ashfield <bruce.ashfield@gmail.com> >>>>>>>> Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org> >>>>>>>> >>>>>>>> Cherry-picked from master 765b13b7305c8d2f222cfc66d77c02e6a088c691 >>>>>>>> >>>>>>>> Signed-off-by: Andreas Helbech Kleist <andreaskleist@gmail.com> >>>>>>>> Signed-off-by: Steve Sakoman <steve@sakoman.com> >>>>>>>> --- >>>>>>>> meta/classes/kernel-arch.bbclass | 7 +++++++ >>>>>>>> meta/classes/kernel.bbclass | 10 ++++++++-- >>>>>>>> 2 files changed, 15 insertions(+), 2 deletions(-) >>>>>>>> >>>>>>>> diff --git a/meta/classes/kernel-arch.bbclass b/meta/classes/kernel-arch.bbclass >>>>>>>> index 4cd08b96fb..0a79dea0af 100644 >>>>>>>> --- a/meta/classes/kernel-arch.bbclass >>>>>>>> +++ b/meta/classes/kernel-arch.bbclass >>>>>>>> @@ -66,3 +66,10 @@ KERNEL_LD = "${CCACHE}${HOST_PREFIX}ld.bfd ${HOST_LD_KERNEL_ARCH}" >>>>>>>> KERNEL_AR = "${CCACHE}${HOST_PREFIX}ar ${HOST_AR_KERNEL_ARCH}" >>>>>>>> TOOLCHAIN ?= "gcc" >>>>>>>> >>>>>>>> +# 6.3+ requires the variable LOCALVERSION to be set to not get a "+" in >>>>>>>> +# the local version. Having it empty means nothing will be added, and any >>>>>>>> +# value will be appended to the local kernel version. This replaces the >>>>>>>> +# use of .scmversion file for setting a localversion without using >>>>>>>> +# the CONFIG_LOCALVERSION option. >>>>>>>> +KERNEL_LOCALVERSION ??= "" >>>>>>>> +export LOCALVERSION ?= "${KERNEL_LOCALVERSION}" >>>>>>>> diff --git a/meta/classes/kernel.bbclass b/meta/classes/kernel.bbclass >>>>>>>> index 5951347361..940f1a3cf4 100644 >>>>>>>> --- a/meta/classes/kernel.bbclass >>>>>>>> +++ b/meta/classes/kernel.bbclass >>>>>>>> @@ -418,7 +418,7 @@ do_compile_kernelmodules() { >>>>>>>> if (grep -q -i -e '^CONFIG_MODULES=y$' ${B}/.config); then >>>>>>>> oe_runmake -C ${B} ${PARALLEL_MAKE} modules ${KERNEL_EXTRA_ARGS} >>>>>>>> >>>>>>>> - # Module.symvers gets updated during the >>>>>>>> + # Module.symvers gets updated during the >>>>>>>> # building of the kernel modules. We need to >>>>>>>> # update this in the shared workdir since some >>>>>>>> # external kernel modules has a dependency on >>>>>>>> @@ -635,7 +635,13 @@ kernel_do_configure() { >>>>>>>> # $ scripts/setlocalversion . => + >>>>>>>> # $ make kernelversion => 2.6.37 >>>>>>>> # $ make kernelrelease => 2.6.37+ >>>>>>>> - touch ${B}/.scmversion ${S}/.scmversion >>>>>>>> + # See kernel-arch.bbclass for post v6.3 removal of the extra >>>>>>>> + # + in localversion. .scmversion is no longer used, and the >>>>>>>> + # variable LOCALVERSION must be used >>>>>>>> + if [ ! -e ${B}/.scmversion -a ! -e ${S}/.scmversion ]; then >>>>>>>> + echo ${KERNEL_LOCALVERSION} > ${B}/.scmversion >>>>>>>> + echo ${KERNEL_LOCALVERSION} > ${S}/.scmversion >>>>>>>> + fi >>>>>>>> >>>>>>>> if [ "${S}" != "${B}" ] && [ -f "${S}/.config" ] && [ ! -f "${B}/.config" ]; then >>>>>>>> mv "${S}/.config" "${B}/.config" >>>>>>>> >>>>>>>> >>>>>>>> >>>>>>>> >>>>>>>> >>>>>>> >>>>>>> -- >>>>>>> Ryan Eatmon reatmon@ti.com >>>>>>> ----------------------------------------- >>>>>>> Texas Instruments, Inc. - LCPD - MGTS >>>>> >>>>> -=-=-=-=-=-=-=-=-=-=-=- >>>>> Links: You receive all messages sent to this group. >>>>> View/Reply Online (#195736): https://lists.openembedded.org/g/openembedded-core/message/195736 >>>>> Mute This Topic: https://lists.openembedded.org/mt/104311283/7501392 >>>>> Group Owner: openembedded-core+owner@lists.openembedded.org >>>>> Unsubscribe: https://lists.openembedded.org/g/openembedded-core/unsub [andreaskleist@gmail.com] >>>>> -=-=-=-=-=-=-=-=-=-=-=- >>>>> >>>> >>> >> > -- Ryan Eatmon reatmon@ti.com ----------------------------------------- Texas Instruments, Inc. - LCPD - MGTS ^ permalink raw reply [flat|nested] 33+ messages in thread
* Re: [OE-core][kirkstone 7/8] kernel: fix localversion in v6.3+ 2024-02-16 8:11 ` Andreas Helbech Kleist 2024-02-16 13:51 ` Andreas Helbech Kleist @ 2024-02-21 14:39 ` Ryan Eatmon 1 sibling, 0 replies; 33+ messages in thread From: Ryan Eatmon @ 2024-02-21 14:39 UTC (permalink / raw) To: Andreas Helbech Kleist, Steve Sakoman; +Cc: openembedded-core On 2/16/2024 2:11 AM, Andreas Helbech Kleist wrote: > On Thu, 2024-02-15 at 19:45 -0600, Ryan Eatmon via > lists.openembedded.org wrote: >> >> On 2/15/2024 7:43 PM, Steve Sakoman wrote: >>> On Thu, Feb 15, 2024 at 3:20 PM Ryan Eatmon <reatmon@ti.com> wrote: >>>> >>>> >>>> With this patch in place we are seeing a breakage on our kernel builds. >>>> This patch was cherry picked from master, but the next patch on this >>>> file in master removes the LOCALVERSION setting... >>>> >>>> https://git.openembedded.org/openembedded-core/commit/meta/classes-recipe/kernel-arch.bbclass?h=master-next&id=b378eec156998eea55ba61e59103cb34fab0d07c >>>> >>>> >>>> There is a disconnect here. Why did we change kirkstone, a stable LTS >>>> version with a partial patch series based on master? >>> >>> The original kirkstone backport request for this patch was sent to the >>> list for review on February 9. >>> >>> There were no comments or objections, so I added it to the patch test >>> queue. No issues were encountered on the autobuilder, so I sent this >>> patch (along with the rest of the patch queue) to the list for a >>> second review opportunity on February 12. Once again there were no >>> comments or objections, so it was merged today February 15. >>> >>> I follow this process because I'm not smart enough, nor do I have time >>> enough to thoroughly research all implications of every patch. I rely >>> on autobuilder testing and community review to minimize breakage. >>> >>> This normally works quite well, but this time it didn't. I'll revert >>> this patch and if someone would like to resubmit a proper series to >>> deal with the issue it is trying to fix I will consider it and run it >>> through the same process. >> >> I can appreciate that. My comment was more for the person who sent the >> patch in the first place. You are doing a great job. I will try and >> pay attention for a replacement patch coming in and review as well. > > I'm sorry about that. I need to use a 6.6 kernel on kirkstone, and this > patch fixed my usecase. It looked like a simple bugfix, and I was not > aware that there was a later fix for this change. > > Is there any way I can reproduce the error you saw Ryan? > > And would it make sense to add an automated test for this usecase? > > I'll test my usecase with both patches backported and re-submit. > > /Andreas > > PS: This is my first submission to oe-core, so I'm just starting to > learn. I have no idea what or how the autobuilder tests, but I'm > willing to learn. The meta-ti layer kernel recipes set KERNEL_LOCALVERSION. With this change in place the modules directory was being name incorrectly which means that our boot testing failed. For example, if we set: KERNEL_LOCALVERSION="xxxx" Then the directory was created: /lib/modules-xxxx-xxxx One too many xxxx in the name. We never saw this issue on master in our testing because they patched the file and backed it out a few days later. >> >> >>> Steve >>> >>> >>>> On 2/12/2024 7:54 AM, Steve Sakoman wrote: >>>>> From: Bruce Ashfield <bruce.ashfield@gmail.com> >>>>> >>>>> During testing of the v6.4 reference kernel, it was noticed that >>>>> on-target modules no longer matched the magic value of the running >>>>> kernel. >>>>> >>>>> This was due to a different localversion in the cross built kernel >>>>> and the scripts / resources created on target. >>>>> >>>>> This was due to changes in the setlocalversion script introduced >>>>> in the v6.3 series. >>>>> >>>>> The .scmversion file is no longer used (or packaged) to inhibit >>>>> the addition of a "+" (through querying of the git status of the >>>>> kernel) or the setting of a local version. >>>>> >>>>> We recently introduced the KERNEL_LOCALVERSION variable to allow >>>>> recipes to place a value in .scmversion, so we extend the use of >>>>> that variable to kernel-arch.bbclass and use it to set the >>>>> exported variable LOCALVERSION. >>>>> >>>>> We must do it at the kernel-arch level, as the variable must be >>>>> exported in any kernel build to ensure that setlocalversion always >>>>> correctly sets the localversion. >>>>> >>>>> Signed-off-by: Bruce Ashfield <bruce.ashfield@gmail.com> >>>>> Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org> >>>>> >>>>> Cherry-picked from master 765b13b7305c8d2f222cfc66d77c02e6a088c691 >>>>> >>>>> Signed-off-by: Andreas Helbech Kleist <andreaskleist@gmail.com> >>>>> Signed-off-by: Steve Sakoman <steve@sakoman.com> >>>>> --- >>>>> meta/classes/kernel-arch.bbclass | 7 +++++++ >>>>> meta/classes/kernel.bbclass | 10 ++++++++-- >>>>> 2 files changed, 15 insertions(+), 2 deletions(-) >>>>> >>>>> diff --git a/meta/classes/kernel-arch.bbclass b/meta/classes/kernel-arch.bbclass >>>>> index 4cd08b96fb..0a79dea0af 100644 >>>>> --- a/meta/classes/kernel-arch.bbclass >>>>> +++ b/meta/classes/kernel-arch.bbclass >>>>> @@ -66,3 +66,10 @@ KERNEL_LD = "${CCACHE}${HOST_PREFIX}ld.bfd ${HOST_LD_KERNEL_ARCH}" >>>>> KERNEL_AR = "${CCACHE}${HOST_PREFIX}ar ${HOST_AR_KERNEL_ARCH}" >>>>> TOOLCHAIN ?= "gcc" >>>>> >>>>> +# 6.3+ requires the variable LOCALVERSION to be set to not get a "+" in >>>>> +# the local version. Having it empty means nothing will be added, and any >>>>> +# value will be appended to the local kernel version. This replaces the >>>>> +# use of .scmversion file for setting a localversion without using >>>>> +# the CONFIG_LOCALVERSION option. >>>>> +KERNEL_LOCALVERSION ??= "" >>>>> +export LOCALVERSION ?= "${KERNEL_LOCALVERSION}" >>>>> diff --git a/meta/classes/kernel.bbclass b/meta/classes/kernel.bbclass >>>>> index 5951347361..940f1a3cf4 100644 >>>>> --- a/meta/classes/kernel.bbclass >>>>> +++ b/meta/classes/kernel.bbclass >>>>> @@ -418,7 +418,7 @@ do_compile_kernelmodules() { >>>>> if (grep -q -i -e '^CONFIG_MODULES=y$' ${B}/.config); then >>>>> oe_runmake -C ${B} ${PARALLEL_MAKE} modules ${KERNEL_EXTRA_ARGS} >>>>> >>>>> - # Module.symvers gets updated during the >>>>> + # Module.symvers gets updated during the >>>>> # building of the kernel modules. We need to >>>>> # update this in the shared workdir since some >>>>> # external kernel modules has a dependency on >>>>> @@ -635,7 +635,13 @@ kernel_do_configure() { >>>>> # $ scripts/setlocalversion . => + >>>>> # $ make kernelversion => 2.6.37 >>>>> # $ make kernelrelease => 2.6.37+ >>>>> - touch ${B}/.scmversion ${S}/.scmversion >>>>> + # See kernel-arch.bbclass for post v6.3 removal of the extra >>>>> + # + in localversion. .scmversion is no longer used, and the >>>>> + # variable LOCALVERSION must be used >>>>> + if [ ! -e ${B}/.scmversion -a ! -e ${S}/.scmversion ]; then >>>>> + echo ${KERNEL_LOCALVERSION} > ${B}/.scmversion >>>>> + echo ${KERNEL_LOCALVERSION} > ${S}/.scmversion >>>>> + fi >>>>> >>>>> if [ "${S}" != "${B}" ] && [ -f "${S}/.config" ] && [ ! -f "${B}/.config" ]; then >>>>> mv "${S}/.config" "${B}/.config" >>>>> >>>>> >>>>> >>>>> >>>>> >>>> >>>> -- >>>> Ryan Eatmon reatmon@ti.com >>>> ----------------------------------------- >>>> Texas Instruments, Inc. - LCPD - MGTS >> >> >> > > > > -=-=-=-=-=-=-=-=-=-=-=- > Links: You receive all messages sent to this group. > View/Reply Online (#195740): https://lists.openembedded.org/g/openembedded-core/message/195740 > Mute This Topic: https://lists.openembedded.org/mt/104311283/6551054 > Group Owner: openembedded-core+owner@lists.openembedded.org > Unsubscribe: https://lists.openembedded.org/g/openembedded-core/unsub [reatmon@ti.com] > -=-=-=-=-=-=-=-=-=-=-=- > -- Ryan Eatmon reatmon@ti.com ----------------------------------------- Texas Instruments, Inc. - LCPD - MGTS ^ permalink raw reply [flat|nested] 33+ messages in thread
* [OE-core][kirkstone 8/8] ghostscript: correct LICENSE with AGPLv3 2024-02-12 13:54 [OE-core][kirkstone 0/8] Patch review Steve Sakoman ` (6 preceding siblings ...) 2024-02-12 13:54 ` [OE-core][kirkstone 7/8] kernel: fix localversion in v6.3+ Steve Sakoman @ 2024-02-12 13:54 ` Steve Sakoman 7 siblings, 0 replies; 33+ messages in thread From: Steve Sakoman @ 2024-02-12 13:54 UTC (permalink / raw) To: openembedded-core From: Kai Kang <kai.kang@windriver.com> The license of ghostscript has switched to Affero GPL since version 9.07 via commit: * 3cc5318 Switch Ghostscript/GhostPDL to Affero GPL https://github.com/ArtifexSoftware/ghostpdl/commit/3cc5318 Correct it with `AGPL-3.0-or-later`. Signed-off-by: Kai Kang <kai.kang@windriver.com> Signed-off-by: Steve Sakoman <steve@sakoman.com> --- meta/recipes-extended/ghostscript/ghostscript_9.55.0.bb | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/meta/recipes-extended/ghostscript/ghostscript_9.55.0.bb b/meta/recipes-extended/ghostscript/ghostscript_9.55.0.bb index e0d1e4618f..e99c740685 100644 --- a/meta/recipes-extended/ghostscript/ghostscript_9.55.0.bb +++ b/meta/recipes-extended/ghostscript/ghostscript_9.55.0.bb @@ -10,7 +10,7 @@ dot-matrix, inkjet and laser models. \ HOMEPAGE = "http://www.ghostscript.com" SECTION = "console/utils" -LICENSE = "GPL-3.0-only" +LICENSE = "AGPL-3.0-or-later" LIC_FILES_CHKSUM = "file://LICENSE;md5=f98ffa763e50cded76f49bce73aade16" DEPENDS = "ghostscript-native tiff jpeg fontconfig cups libpng" -- 2.34.1 ^ permalink raw reply related [flat|nested] 33+ messages in thread
* [OE-core][kirkstone 0/8] Patch review
@ 2025-11-03 20:59 Steve Sakoman
0 siblings, 0 replies; 33+ messages in thread
From: Steve Sakoman @ 2025-11-03 20:59 UTC (permalink / raw)
To: openembedded-core
Please review this set of changes for kirkstone and have comments back by
end of day Wednesday, November 5
Passed a-full on autobuilder:
https://autobuilder.yoctoproject.org/valkyrie/#/builders/29/builds/2677
The following changes since commit 99204008786f659ab03538cd2ae2fd23ed4164c5:
build-appliance-image: Update to kirkstone head revision (2025-10-31 06:30:23 -0700)
are available in the Git repository at:
https://git.openembedded.org/openembedded-core-contrib stable/kirkstone-nut
https://git.openembedded.org/openembedded-core-contrib/log/?h=stable/kirkstone-nut
Archana Polampalli (1):
openssh: fix CVE-2025-61985
Hitendra Prajapati (1):
go: fix CVE-2024-24783
Hongxu Jia (1):
u-boot: fix CVE-2024-42040
Jason Schonberg (1):
Don't use ftp.gnome.org
Peter Marko (3):
wpa-supplicant: patch CVE-2025-24912
binutils: patch CVE-2025-11412
binutils: patch CVE-2025-11413
Praveen Kumar (1):
bind: upgrade 9.18.33 -> 9.18.41
.../u-boot/files/CVE-2024-42040.patch | 56 +++++++++++++
meta/recipes-bsp/u-boot/u-boot-common.inc | 4 +-
.../bind/{bind_9.18.33.bb => bind_9.18.41.bb} | 2 +-
.../openssh/openssh/CVE-2025-61985.patch | 35 ++++++++
.../openssh/openssh_8.9p1.bb | 1 +
.../wpa-supplicant/CVE-2025-24912-01.patch | 79 ++++++++++++++++++
.../wpa-supplicant/CVE-2025-24912-02.patch | 70 ++++++++++++++++
.../wpa-supplicant/wpa-supplicant_2.10.bb | 2 +
.../binutils/binutils-2.38.inc | 2 +
.../binutils/binutils/CVE-2025-11412.patch | 35 ++++++++
.../binutils/binutils/CVE-2025-11413.patch | 38 +++++++++
meta/recipes-devtools/go/go-1.17.13.inc | 1 +
.../go/go-1.21/CVE-2024-24783.patch | 83 +++++++++++++++++++
.../python/python3-pygobject_3.42.0.bb | 2 +-
meta/recipes-devtools/vala/vala.inc | 2 +-
meta/recipes-gnome/gtk+/gtk+3_3.24.34.bb | 2 +-
meta/recipes-gnome/libgudev/libgudev_237.bb | 2 +-
.../recipes-support/libxslt/libxslt_1.1.35.bb | 2 +-
18 files changed, 411 insertions(+), 7 deletions(-)
create mode 100644 meta/recipes-bsp/u-boot/files/CVE-2024-42040.patch
rename meta/recipes-connectivity/bind/{bind_9.18.33.bb => bind_9.18.41.bb} (97%)
create mode 100644 meta/recipes-connectivity/openssh/openssh/CVE-2025-61985.patch
create mode 100644 meta/recipes-connectivity/wpa-supplicant/wpa-supplicant/CVE-2025-24912-01.patch
create mode 100644 meta/recipes-connectivity/wpa-supplicant/wpa-supplicant/CVE-2025-24912-02.patch
create mode 100644 meta/recipes-devtools/binutils/binutils/CVE-2025-11412.patch
create mode 100644 meta/recipes-devtools/binutils/binutils/CVE-2025-11413.patch
create mode 100644 meta/recipes-devtools/go/go-1.21/CVE-2024-24783.patch
--
2.43.0
^ permalink raw reply [flat|nested] 33+ messages in thread* [OE-core][kirkstone 0/8] Patch review
@ 2025-10-17 20:43 Steve Sakoman
0 siblings, 0 replies; 33+ messages in thread
From: Steve Sakoman @ 2025-10-17 20:43 UTC (permalink / raw)
To: openembedded-core
Please review this set of changes for kirkstone and have comments back by
end of day Tuesday, October 21
Passed a-full on autobuilder:
https://autobuilder.yoctoproject.org/valkyrie/#/builders/29/builds/2607
The following changes since commit 8f1000d9dad5e51f08a40b0f6650204425cc8efb:
glibc: : PTHREAD_COND_INITIALIZER compatibility with pre-2.41 versions (bug 32786) (2025-10-14 10:35:12 -0700)
are available in the Git repository at:
https://git.openembedded.org/openembedded-core-contrib stable/kirkstone-nut
https://git.openembedded.org/openembedded-core-contrib/log/?h=stable/kirkstone-nut
Bruce Ashfield (4):
linux-yocto/5.15: update to v5.15.188
linux-yocto/5.15: update to v5.15.189
linux-yocto/5.15: update to v5.15.193
linux-yocto/5.15: update to v5.15.194
Peter Marko (1):
python3: upgrade 3.10.18 -> 3.10.19
Rajeshkumar Ramasamy (2):
glib-networking: fix CVE-2025-60018
glib-networking: fix CVE-2025-60019
Saravanan (1):
cmake: fix CVE-2025-9301
.../glib-networking/CVE-2025-60018.patch | 83 +++++++
.../glib-networking/CVE-2025-60019.patch | 137 +++++++++++
.../glib-networking/glib-networking_2.72.2.bb | 2 +
.../cmake/cmake/CVE-2025-9301.patch | 71 ++++++
meta/recipes-devtools/cmake/cmake_3.22.3.bb | 1 +
...e-treat-overflow-in-UID-GID-as-failu.patch | 2 +-
.../python/python3/CVE-2025-8194.patch | 219 ------------------
...{python3_3.10.18.bb => python3_3.10.19.bb} | 3 +-
.../linux/linux-yocto-rt_5.15.bb | 6 +-
.../linux/linux-yocto-tiny_5.15.bb | 6 +-
meta/recipes-kernel/linux/linux-yocto_5.15.bb | 26 +--
11 files changed, 315 insertions(+), 241 deletions(-)
create mode 100644 meta/recipes-core/glib-networking/glib-networking/CVE-2025-60018.patch
create mode 100644 meta/recipes-core/glib-networking/glib-networking/CVE-2025-60019.patch
create mode 100644 meta/recipes-devtools/cmake/cmake/CVE-2025-9301.patch
delete mode 100644 meta/recipes-devtools/python/python3/CVE-2025-8194.patch
rename meta/recipes-devtools/python/{python3_3.10.18.bb => python3_3.10.19.bb} (99%)
--
2.43.0
^ permalink raw reply [flat|nested] 33+ messages in thread* [OE-core][kirkstone 0/8] Patch review @ 2025-08-13 21:28 Steve Sakoman 0 siblings, 0 replies; 33+ messages in thread From: Steve Sakoman @ 2025-08-13 21:28 UTC (permalink / raw) To: openembedded-core Please review this set of changes for kirkstone and have comments back by end of day Friday, August 15 Passed a-full on autobuilder: https://autobuilder.yoctoproject.org/valkyrie/#/builders/29/builds/2196 The following changes since commit bd620eb14660075fd0f7476bbbb65d5da6293874: build-appliance-image: Update to kirkstone head revision (2025-08-08 06:31:30 -0700) are available in the Git repository at: https://git.openembedded.org/openembedded-core-contrib stable/kirkstone-nut https://git.openembedded.org/openembedded-core-contrib/log/?h=stable/kirkstone-nut Guocai He (1): gnupg: disable tests to avoid running target binaries at build time Hitendra Prajapati (1): libxslt: fix CVE-2023-40403 Peter Marko (3): python3: patch CVE-2025-8194 go: ignore CVE-2025-0913 libarchive: patch CVE-2025-5918 Quentin Schulz (1): go-helloworld: fix license Yogita Urade (2): tiff: fix CVE-2025-8176 tiff: fix CVE-2025-8177 meta/recipes-devtools/go/go-1.17.13.inc | 2 +- .../python/python3/CVE-2025-8194.patch | 219 +++++++++++ .../python/python3_3.10.18.bb | 7 +- .../go-examples/go-helloworld_0.1.bb | 4 +- .../0001-FILE-seeking-support-2539.patch | 190 ++++++++++ .../0001-Improve-lseek-handling-2564.patch | 320 ++++++++++++++++ .../libarchive/libarchive/CVE-2025-5918.patch | 217 +++++++++++ .../libarchive/libarchive_3.6.2.bb | 3 + .../libtiff/tiff/CVE-2025-8176-0001.patch | 61 +++ .../libtiff/tiff/CVE-2025-8176-0002.patch | 31 ++ .../libtiff/tiff/CVE-2025-8176-0003.patch | 28 ++ .../libtiff/tiff/CVE-2025-8177.patch | 35 ++ meta/recipes-multimedia/libtiff/tiff_4.3.0.bb | 4 + meta/recipes-support/gnupg/gnupg_2.3.7.bb | 1 + .../libxslt/libxslt/CVE-2023-40403-001.patch | 257 +++++++++++++ .../libxslt/libxslt/CVE-2023-40403-002.patch | 147 ++++++++ .../libxslt/libxslt/CVE-2023-40403-003.patch | 231 ++++++++++++ .../libxslt/libxslt/CVE-2023-40403-004.patch | 349 ++++++++++++++++++ .../libxslt/libxslt/CVE-2023-40403-005.patch | 55 +++ .../recipes-support/libxslt/libxslt_1.1.35.bb | 5 + 20 files changed, 2160 insertions(+), 6 deletions(-) create mode 100644 meta/recipes-devtools/python/python3/CVE-2025-8194.patch create mode 100644 meta/recipes-extended/libarchive/libarchive/0001-FILE-seeking-support-2539.patch create mode 100644 meta/recipes-extended/libarchive/libarchive/0001-Improve-lseek-handling-2564.patch create mode 100644 meta/recipes-extended/libarchive/libarchive/CVE-2025-5918.patch create mode 100644 meta/recipes-multimedia/libtiff/tiff/CVE-2025-8176-0001.patch create mode 100644 meta/recipes-multimedia/libtiff/tiff/CVE-2025-8176-0002.patch create mode 100644 meta/recipes-multimedia/libtiff/tiff/CVE-2025-8176-0003.patch create mode 100644 meta/recipes-multimedia/libtiff/tiff/CVE-2025-8177.patch create mode 100644 meta/recipes-support/libxslt/libxslt/CVE-2023-40403-001.patch create mode 100644 meta/recipes-support/libxslt/libxslt/CVE-2023-40403-002.patch create mode 100644 meta/recipes-support/libxslt/libxslt/CVE-2023-40403-003.patch create mode 100644 meta/recipes-support/libxslt/libxslt/CVE-2023-40403-004.patch create mode 100644 meta/recipes-support/libxslt/libxslt/CVE-2023-40403-005.patch -- 2.43.0 ^ permalink raw reply [flat|nested] 33+ messages in thread
* [OE-core][kirkstone 0/8] Patch review @ 2025-04-01 22:36 Steve Sakoman 0 siblings, 0 replies; 33+ messages in thread From: Steve Sakoman @ 2025-04-01 22:36 UTC (permalink / raw) To: openembedded-core Please review this set of changes for kirkstone and have comments back by end of day Thursday, April 3 Passed a-full on autobuilder: https://autobuilder.yoctoproject.org/valkyrie/#/builders/29/builds/1311 The following changes since commit 453c5c8d9031be2b3a25e2a04e0f5f6325ef7298: cve-update-nvd2-native: handle missing vulnStatus (2025-03-31 09:13:54 -0700) are available in the Git repository at: https://git.openembedded.org/openembedded-core-contrib stable/kirkstone-nut https://git.openembedded.org/openembedded-core-contrib/log/?h=stable/kirkstone-nut Deepesh Varatharajan (1): llvm : Fix CVE-2024-0151 Divya Chellam (1): zlib: fix CVE-2014-9485 Guocai He (1): mesa: Update SRC_URI Haixiao Yan (1): glibc: Add single-threaded fast path to rand() Michael Halstead (1): yocto-uninative: Update to 4.7 for glibc 2.41 Peter Marko (3): libarchive: ignore CVE-2025-1632 perl: ignore CVE-2023-47038 freetype: patch CVE-2025-27363 meta/conf/distro/include/yocto-uninative.inc | 10 +- ...dd-single-threaded-fast-path-to-rand.patch | 47 + meta/recipes-core/glibc/glibc_2.35.bb | 1 + .../zlib/zlib/CVE-2014-9485.patch | 64 + meta/recipes-core/zlib/zlib_1.2.11.bb | 1 + .../llvm/llvm/CVE-2024-0151.patch | 1087 +++++++++++++++++ meta/recipes-devtools/llvm/llvm_git.bb | 1 + meta/recipes-devtools/perl/perl_5.34.3.bb | 2 + .../libarchive/libarchive_3.6.2.bb | 2 + .../freetype/freetype/CVE-2025-27363.patch | 44 + .../freetype/freetype_2.11.1.bb | 1 + .../recipes-graphics/mesa/mesa-demos_8.4.0.bb | 2 +- meta/recipes-graphics/mesa/mesa.inc | 2 +- 13 files changed, 1257 insertions(+), 7 deletions(-) create mode 100644 meta/recipes-core/glibc/glibc/0001-stdlib-Add-single-threaded-fast-path-to-rand.patch create mode 100644 meta/recipes-core/zlib/zlib/CVE-2014-9485.patch create mode 100644 meta/recipes-devtools/llvm/llvm/CVE-2024-0151.patch create mode 100644 meta/recipes-graphics/freetype/freetype/CVE-2025-27363.patch -- 2.43.0 ^ permalink raw reply [flat|nested] 33+ messages in thread
* [OE-core][kirkstone 0/8] Patch review
@ 2024-08-04 17:08 Steve Sakoman
0 siblings, 0 replies; 33+ messages in thread
From: Steve Sakoman @ 2024-08-04 17:08 UTC (permalink / raw)
To: openembedded-core
Please review this set of changes for kirkstone and have comments back by
end of day Tuesday, August 6
Passed a-full on autobuilder:
https://autobuilder.yoctoproject.org/typhoon/#/builders/83/builds/7206
The following changes since commit c6cafd2aa50357c80fbab79741d575ff567c5766:
gcc-runtime: remove bashism (2024-07-31 04:59:21 -0700)
are available in the Git repository at:
https://git.openembedded.org/openembedded-core-contrib stable/kirkstone-nut
https://git.openembedded.org/openembedded-core-contrib/log/?h=stable/kirkstone-nut
Ashish Sharma (1):
gtk+3 : backport fix for CVE-2024-6655
Bruce Ashfield (5):
linux-yocto/5.15: update to v5.15.158
linux-yocto/5.15: update to v5.15.160
linux-yocto/5.15: update to v5.15.161
linux-yocto/5.15: update to v5.15.162
linux-yocto/5.15: update to v5.15.164
Siddharth Doshi (1):
lttng-modules: Upgrade 2.13.9 -> 2.13.14
Soumya Sambu (1):
go: Fix CVE-2024-24789
meta/recipes-devtools/go/go-1.17.13.inc | 1 +
.../go/go-1.21/CVE-2024-24789.patch | 78 +++++++++++++++++++
.../gtk+/gtk+3/CVE-2024-6655.patch | 39 ++++++++++
meta/recipes-gnome/gtk+/gtk+3_3.24.34.bb | 1 +
.../linux/linux-yocto-rt_5.15.bb | 6 +-
.../linux/linux-yocto-tiny_5.15.bb | 6 +-
meta/recipes-kernel/linux/linux-yocto_5.15.bb | 26 +++----
.../0009-Rename-genhd-wrapper-to-blkdev.patch | 19 +++--
...les_2.13.9.bb => lttng-modules_2.13.14.bb} | 4 +-
9 files changed, 151 insertions(+), 29 deletions(-)
create mode 100644 meta/recipes-devtools/go/go-1.21/CVE-2024-24789.patch
create mode 100644 meta/recipes-gnome/gtk+/gtk+3/CVE-2024-6655.patch
rename meta/recipes-kernel/lttng/{lttng-modules_2.13.9.bb => lttng-modules_2.13.14.bb} (89%)
--
2.34.1
^ permalink raw reply [flat|nested] 33+ messages in thread* [OE-core][kirkstone 0/8] Patch review
@ 2024-05-24 12:14 Steve Sakoman
0 siblings, 0 replies; 33+ messages in thread
From: Steve Sakoman @ 2024-05-24 12:14 UTC (permalink / raw)
To: openembedded-core
Please review this set of changes for kirkstone and have comments back by
end of day Tuesday, May 28
Passed a-full on autobuilder:
https://autobuilder.yoctoproject.org/typhoon/#/builders/83/builds/6956
The following changes since commit f85d5dfc91d536a00669ca3148d8c3b2727b183d:
libpciaccess: Remove duplicated license entry (2024-05-10 05:05:54 -0700)
are available in the Git repository at:
https://git.openembedded.org/openembedded-core-contrib stable/kirkstone-nut
https://git.openembedded.org/openembedded-core-contrib/log/?h=stable/kirkstone-nut
Bob Henz (1):
systemd-systemctl: Fix WantedBy processing
Colin McAllister (1):
initscripts: Add custom mount args for /var/lib
Dmitry Baryshkov (1):
go.bbclass: fix path to linker in native Go builds
Joerg Vehlow (1):
go: Always pass interpreter to linker
Peter Marko (1):
openssl: patch CVE-2024-4603
Stefan Herbrechtsmeier (1):
classes: go-mod: do not pack go mod cache
Vijay Anusuri (1):
binutils: Rename CVE-2022-38126 patch to CVE-2022-35205
Yogita Urade (1):
libarchive: fix CVE-2024-26256
meta/classes/go-mod.bbclass | 4 +
meta/classes/go.bbclass | 6 +-
.../openssl/openssl/CVE-2024-4603.patch | 180 ++++++++++++++++++
.../openssl/openssl_3.0.13.bb | 1 +
.../initscripts-1.0/read-only-rootfs-hook.sh | 4 +-
.../initscripts/initscripts_1.0.bb | 2 +
.../systemd/systemd-systemctl/systemctl | 11 ++
.../binutils/binutils-2.38.inc | 2 +-
...-38126.patch => 0016-CVE-2022-35205.patch} | 3 +-
.../libarchive/CVE-2024-26256.patch | 29 +++
.../libarchive/libarchive_3.6.2.bb | 5 +-
11 files changed, 240 insertions(+), 7 deletions(-)
create mode 100644 meta/recipes-connectivity/openssl/openssl/CVE-2024-4603.patch
rename meta/recipes-devtools/binutils/binutils/{0016-CVE-2022-38126.patch => 0016-CVE-2022-35205.patch} (94%)
create mode 100644 meta/recipes-extended/libarchive/libarchive/CVE-2024-26256.patch
--
2.34.1
^ permalink raw reply [flat|nested] 33+ messages in thread* [OE-core][kirkstone 0/8] Patch review @ 2024-03-12 13:53 Steve Sakoman 0 siblings, 0 replies; 33+ messages in thread From: Steve Sakoman @ 2024-03-12 13:53 UTC (permalink / raw) To: openembedded-core Please review this set of changes for kirkstone and have comments back by end of day Wednesday, March 13 Passed a-full on autobuilder: https://autobuilder.yoctoproject.org/typhoon/#/builders/83/builds/6670 The following changes since commit e5aae8a371717215a7d78459788ad67dfaefe37e: golang: Fix CVE-2023-45289 & CVE-2023-45290 (2024-03-07 04:18:33 -1000) are available in the Git repository at: https://git.openembedded.org/openembedded-core-contrib stable/kirkstone-nut https://git.openembedded.org/openembedded-core-contrib/log/?h=stable/kirkstone-nut Bruce Ashfield (6): linux-yocto/5.15: update to v5.15.149 linux-yocto/5.15: update CVE exclusions linux-yocto/5.10: update to v5.10.210 linux-yocto/5.15: update to v5.15.150 linux-yocto/5.15: update CVE exclusions (5.15.150) linux-yocto/5.15: fix partion scanning Nikhil R (1): librsvg: Fix do_package_qa error for librsvg Vivek Kumbhar (1): go: Backport fix CVE-2024-24784 & CVE-2024-24785 meta/recipes-devtools/go/go-1.17.13.inc | 2 + .../go/go-1.18/CVE-2024-24784.patch | 207 ++++++++++++++++++ .../go/go-1.18/CVE-2024-24785.patch | 196 +++++++++++++++++ meta/recipes-gnome/librsvg/librsvg_2.52.10.bb | 2 + .../linux/cve-exclusion_5.15.inc | 197 ++++++++++++++++- .../linux/linux-yocto-rt_5.10.bb | 4 +- .../linux/linux-yocto-rt_5.15.bb | 6 +- .../linux/linux-yocto-tiny_5.10.bb | 6 +- .../linux/linux-yocto-tiny_5.15.bb | 6 +- meta/recipes-kernel/linux/linux-yocto_5.10.bb | 22 +- meta/recipes-kernel/linux/linux-yocto_5.15.bb | 26 +-- 11 files changed, 632 insertions(+), 42 deletions(-) create mode 100644 meta/recipes-devtools/go/go-1.18/CVE-2024-24784.patch create mode 100644 meta/recipes-devtools/go/go-1.18/CVE-2024-24785.patch -- 2.34.1 ^ permalink raw reply [flat|nested] 33+ messages in thread
* [OE-core][kirkstone 0/8] Patch review
@ 2023-11-29 23:04 Steve Sakoman
0 siblings, 0 replies; 33+ messages in thread
From: Steve Sakoman @ 2023-11-29 23:04 UTC (permalink / raw)
To: openembedded-core
Please review this set of changes for kirkstone and have comments back by
end of day Friday, December 1
Passed a-full on autobuilder:
https://autobuilder.yoctoproject.org/typhoon/#/builders/83/builds/6260
The following changes since commit 8726ae02d760270f9e7fe7ef5715d8f7553371ce:
goarch: Move Go architecture mapping to a library (2023-11-21 05:32:39 -1000)
are available in the Git repository at:
https://git.openembedded.org/openembedded-core-contrib stable/kirkstone-nut
https://git.openembedded.org/openembedded-core-contrib/log/?h=stable/kirkstone-nut
Archana Polampalli (2):
gstreamer1.0-plugins-bad: fix CVE-2023-44429
vim: Upgrade 9.0.2048 -> 9.0.2068
Hitendra Prajapati (1):
grub: fix CVE-2023-4693
Li Wang (1):
systemtap_git: fix used uninitialized error
Ninad Palsule (1):
kernel-fitImage: Strip path component from dtb
Richard Purdie (1):
vim: Improve locale handling
Steve Sakoman (1):
vim: use upstream generated .po files
Vivek Kumbhar (1):
openssl: fix CVE-2023-5678 Generating excessively long X9.42 DH keys
or checking excessively long X9.42 DH keys or parameters may be very
slow
meta/classes/kernel-fitimage.bbclass | 5 +
.../grub/files/CVE-2023-4693.patch | 62 ++++++
meta/recipes-bsp/grub/grub2.inc | 1 +
.../openssl/openssl/CVE-2023-5678.patch | 180 ++++++++++++++++++
.../openssl/openssl_3.0.12.bb | 1 +
...x-Prevent-Werror-maybe-uninitialized.patch | 53 ++++++
.../recipes-kernel/systemtap/systemtap_git.bb | 1 +
.../CVE-2023-44429.patch | 38 ++++
.../gstreamer1.0-plugins-bad_1.20.7.bb | 1 +
meta/recipes-support/vim/vim.inc | 20 +-
10 files changed, 350 insertions(+), 12 deletions(-)
create mode 100644 meta/recipes-bsp/grub/files/CVE-2023-4693.patch
create mode 100644 meta/recipes-connectivity/openssl/openssl/CVE-2023-5678.patch
create mode 100644 meta/recipes-kernel/systemtap/systemtap/0001-bpf-translate.cxx-Prevent-Werror-maybe-uninitialized.patch
create mode 100644 meta/recipes-multimedia/gstreamer/gstreamer1.0-plugins-bad/CVE-2023-44429.patch
--
2.34.1
^ permalink raw reply [flat|nested] 33+ messages in thread* [OE-core][kirkstone 0/8] Patch review @ 2023-09-13 14:30 Steve Sakoman 0 siblings, 0 replies; 33+ messages in thread From: Steve Sakoman @ 2023-09-13 14:30 UTC (permalink / raw) To: openembedded-core Please review this set of changes for kirkstone and have comments back by end of day Friday, September 15. Passed a-full on autobuilder: https://autobuilder.yoctoproject.org/typhoon/#/builders/83/builds/5876 The following changes since commit 47a1dd7f389e3cf4ac2dc5fc21dccc870aafab4a: sysklogd: fix integration with systemd-journald (2023-09-05 13:34:12 -1000) are available in the Git repository at: https://git.openembedded.org/openembedded-core-contrib stable/kirkstone-nut http://cgit.openembedded.org/openembedded-core-contrib/log/?h=stable/kirkstone-nut Meenali Gupta (1): flac: fix CVE-2020-22219 Michael Halstead (1): yocto-uninative: Update to 4.3 Narpat Mali (1): python3-pygments: Fix CVE-2022-40896 Siddharth Doshi (1): gdb: Fix CVE-2023-39128 Soumya Sambu (1): libxml2: Fix CVE-2023-39615 Yogita Urade (3): dropbear: fix CVE-2023-36328 qemu: fix CVE-2021-3638 webkitgtk: fix CVE-2022-48503 meta/conf/distro/include/yocto-uninative.inc | 8 +- meta/recipes-core/dropbear/dropbear.inc | 1 + .../dropbear/dropbear/CVE-2023-36328.patch | 144 +++++++++++ .../libxml/libxml2/CVE-2023-39615-0001.patch | 37 +++ .../libxml/libxml2/CVE-2023-39615-0002.patch | 72 ++++++ meta/recipes-core/libxml/libxml2_2.9.14.bb | 2 + meta/recipes-devtools/gdb/gdb.inc | 1 + .../gdb/gdb/0011-CVE-2023-39128.patch | 75 ++++++ .../python3-pygments/CVE-2022-40896.patch | 124 ++++++++++ .../python/python3-pygments_2.11.2.bb | 2 + meta/recipes-devtools/qemu/qemu.inc | 1 + .../qemu/qemu/CVE-2021-3638.patch | 88 +++++++ .../flac/files/CVE-2020-22219.patch | 197 +++++++++++++++ meta/recipes-multimedia/flac/flac_1.3.4.bb | 1 + .../webkit/webkitgtk/CVE-2022-48503.patch | 225 ++++++++++++++++++ meta/recipes-sato/webkit/webkitgtk_2.36.8.bb | 1 + 16 files changed, 975 insertions(+), 4 deletions(-) create mode 100644 meta/recipes-core/dropbear/dropbear/CVE-2023-36328.patch create mode 100644 meta/recipes-core/libxml/libxml2/CVE-2023-39615-0001.patch create mode 100644 meta/recipes-core/libxml/libxml2/CVE-2023-39615-0002.patch create mode 100644 meta/recipes-devtools/gdb/gdb/0011-CVE-2023-39128.patch create mode 100644 meta/recipes-devtools/python/python3-pygments/CVE-2022-40896.patch create mode 100644 meta/recipes-devtools/qemu/qemu/CVE-2021-3638.patch create mode 100644 meta/recipes-multimedia/flac/files/CVE-2020-22219.patch create mode 100644 meta/recipes-sato/webkit/webkitgtk/CVE-2022-48503.patch -- 2.34.1 ^ permalink raw reply [flat|nested] 33+ messages in thread
* [OE-core][kirkstone 0/8] Patch review
@ 2023-08-27 20:52 Steve Sakoman
0 siblings, 0 replies; 33+ messages in thread
From: Steve Sakoman @ 2023-08-27 20:52 UTC (permalink / raw)
To: openembedded-core
Please review this set of changes for kirkstone and have comments back by
end of day Tuesday, August 29.
Passed a-full on autobuilder:
https://autobuilder.yoctoproject.org/typhoon/#/builders/83/builds/5789
The following changes since commit ea920e3c8075f3a1b79039341f8c889f6197a07f:
glibc-locale: use stricter matching for metapackages' runtime dependencies (2023-08-22 07:07:13 -1000)
are available in the Git repository at:
https://git.openembedded.org/openembedded-core-contrib stable/kirkstone-nut
http://cgit.openembedded.org/openembedded-core-contrib/log/?h=stable/kirkstone-nut
Narpat Mali (2):
ffmpeg: add CVE_CHECK_IGNORE for CVE-2023-39018
python3-git: upgrade 3.1.27 -> 3.1.32
Ross Burton (3):
linux-yocto: add script to generate kernel CVE_CHECK_IGNORE entries
linux/cve-exclusion: add generated CVE_CHECK_IGNORES.
linux/cve-exclusion: remove obsolete manual entries
Siddharth (1):
Qemu: Resolve undefined reference issue in CVE-2023-2861
Soumya Sambu (1):
go: Fix CVE-2023-29409
Yogita Urade (1):
nghttp2: fix CVE-2023-35945
meta/recipes-devtools/go/go-1.17.13.inc | 1 +
.../go/go-1.19/CVE-2023-29409.patch | 175 +
...-git-CVE-2022-24439-fix-from-PR-1518.patch | 97 -
...-git-CVE-2022-24439-fix-from-PR-1521.patch | 488 --
...n3-git_3.1.27.bb => python3-git_3.1.32.bb} | 6 +-
.../qemu/qemu/CVE-2023-2861.patch | 66 +-
meta/recipes-kernel/linux/cve-exclusion.inc | 869 --
.../linux/cve-exclusion_5.15.inc | 7193 +++++++++++++++++
.../linux/generate-cve-exclusions.py | 101 +
meta/recipes-kernel/linux/linux-yocto_5.15.bb | 1 +
.../recipes-multimedia/ffmpeg/ffmpeg_5.0.1.bb | 6 +
.../nghttp2/nghttp2/CVE-2023-35945.patch | 151 +
.../recipes-support/nghttp2/nghttp2_1.47.0.bb | 1 +
13 files changed, 7667 insertions(+), 1488 deletions(-)
create mode 100644 meta/recipes-devtools/go/go-1.19/CVE-2023-29409.patch
delete mode 100644 meta/recipes-devtools/python/python3-git/0001-python3-git-CVE-2022-24439-fix-from-PR-1518.patch
delete mode 100644 meta/recipes-devtools/python/python3-git/0001-python3-git-CVE-2022-24439-fix-from-PR-1521.patch
rename meta/recipes-devtools/python/{python3-git_3.1.27.bb => python3-git_3.1.32.bb} (80%)
create mode 100644 meta/recipes-kernel/linux/cve-exclusion_5.15.inc
create mode 100755 meta/recipes-kernel/linux/generate-cve-exclusions.py
create mode 100644 meta/recipes-support/nghttp2/nghttp2/CVE-2023-35945.patch
--
2.34.1
^ permalink raw reply [flat|nested] 33+ messages in thread* [OE-core][kirkstone 0/8] Patch review
@ 2023-07-24 2:33 Steve Sakoman
0 siblings, 0 replies; 33+ messages in thread
From: Steve Sakoman @ 2023-07-24 2:33 UTC (permalink / raw)
To: openembedded-core
Please review this set of changes for kirkstone and have comments back by
end of day Tuesday.
Passed a-full on autobuilder:
https://autobuilder.yoctoproject.org/typhoon/#/builders/83/builds/5638
The following changes since commit d877d5f07772ec4a05332068ddc03cf387313036:
cmake: Fix CMAKE_SYSTEM_PROCESSOR setting for SDK (2023-07-17 04:45:01 -1000)
are available in the Git repository at:
https://git.openembedded.org/openembedded-core-contrib stable/kirkstone-nut
http://cgit.openembedded.org/openembedded-core-contrib/log/?h=stable/kirkstone-nut
Archana Polampalli (1):
ghostscript: fix CVE-2023-36664
Bruce Ashfield (2):
linux-yocto/5.15: update to v5.15.119
linux-yocto/5.15: update to v5.15.120
Richard Purdie (1):
gcc-testsuite: Fix ppc cpu specification
Ross Burton (2):
gcc: don't pass --enable-standard-branch-protection
machine/arch-arm64: add -mbranch-protection=standard
Vijay Anusuri (1):
qemu: backport Debian patch to fix CVE-2023-0330
Xiangyu Chen (1):
package.bbclass: moving field data process before variable process in
process_pkgconfig
meta/classes/package.bbclass | 12 +-
meta/conf/machine/include/arm/arch-arm64.inc | 5 +
.../gcc/gcc-configure-common.inc | 1 -
meta/recipes-devtools/gcc/gcc-testsuite.inc | 5 +-
meta/recipes-devtools/qemu/qemu.inc | 1 +
.../qemu/qemu/CVE-2023-0330.patch | 75 +++++++++
.../ghostscript/CVE-2023-36664-0001.patch | 146 ++++++++++++++++++
.../ghostscript/CVE-2023-36664-0002.patch | 60 +++++++
.../ghostscript/ghostscript_9.55.0.bb | 2 +
.../linux/linux-yocto-rt_5.15.bb | 6 +-
.../linux/linux-yocto-tiny_5.15.bb | 6 +-
meta/recipes-kernel/linux/linux-yocto_5.15.bb | 26 ++--
12 files changed, 317 insertions(+), 28 deletions(-)
create mode 100644 meta/recipes-devtools/qemu/qemu/CVE-2023-0330.patch
create mode 100644 meta/recipes-extended/ghostscript/ghostscript/CVE-2023-36664-0001.patch
create mode 100644 meta/recipes-extended/ghostscript/ghostscript/CVE-2023-36664-0002.patch
--
2.34.1
^ permalink raw reply [flat|nested] 33+ messages in thread* [OE-core][kirkstone 0/8] Patch review @ 2023-04-22 15:54 Steve Sakoman 0 siblings, 0 replies; 33+ messages in thread From: Steve Sakoman @ 2023-04-22 15:54 UTC (permalink / raw) To: openembedded-core Please review this set of patches for kirkstone and have comments back by end of day Tuesday. Passed a-full on autobuilder: https://autobuilder.yoctoproject.org/typhoon/#/builders/83/builds/5209 The following changes since commit b67e714b367a08fdeeeff68c2d9495ec9bc07304: package.bbclass: correct check for /build in copydebugsources() (2023-04-14 07:19:08 -1000) are available in the Git repository at: https://git.openembedded.org/openembedded-core-contrib stable/kirkstone-nut http://cgit.openembedded.org/openembedded-core-contrib/log/?h=stable/kirkstone-nut Hitendra Prajapati (2): ruby: CVE-2023-28756 ReDoS vulnerability in Time screen: CVE-2023-24626 allows sending SIGHUP to arbitrary PIDs Peter Marko (1): go: ignore CVE-2022-41716 Shubham Kulkarni (1): go-runtime: Security fix for CVE-2022-41722 Siddharth Doshi (1): curl: Security fix for CVE-2023-27535, CVE-2023-27536, CVE-2023-27538 Sundeep KOKKONDA (1): cargo : non vulnerable cve-2022-46176 added to excluded list Vivek Kumbhar (1): go: fix CVE-2023-24537 Infinite loop in parsing Xiangyu Chen (1): shadow: backport patch to fix CVE-2023-29383 .../distro/include/cve-extra-exclusions.inc | 5 + meta/recipes-devtools/go/go-1.17.13.inc | 5 + .../go/go-1.18/CVE-2022-41722.patch | 103 +++++++++ .../go/go-1.18/CVE-2023-24537.patch | 75 +++++++ .../ruby/ruby/CVE-2023-28756.patch | 73 +++++++ meta/recipes-devtools/ruby/ruby_3.1.3.bb | 1 + .../screen/screen/CVE-2023-24626.patch | 40 ++++ meta/recipes-extended/screen/screen_4.9.0.bb | 1 + .../files/0001-Overhaul-valid_field.patch | 65 ++++++ .../shadow/files/CVE-2023-29383.patch | 53 +++++ meta/recipes-extended/shadow/shadow.inc | 2 + .../curl/curl/CVE-2023-27535-pre1.patch | 196 ++++++++++++++++++ .../CVE-2023-27535_and_CVE-2023-27538.patch | 170 +++++++++++++++ .../curl/curl/CVE-2023-27536.patch | 52 +++++ meta/recipes-support/curl/curl_7.82.0.bb | 3 + 15 files changed, 844 insertions(+) create mode 100644 meta/recipes-devtools/go/go-1.18/CVE-2022-41722.patch create mode 100644 meta/recipes-devtools/go/go-1.18/CVE-2023-24537.patch create mode 100644 meta/recipes-devtools/ruby/ruby/CVE-2023-28756.patch create mode 100644 meta/recipes-extended/screen/screen/CVE-2023-24626.patch create mode 100644 meta/recipes-extended/shadow/files/0001-Overhaul-valid_field.patch create mode 100644 meta/recipes-extended/shadow/files/CVE-2023-29383.patch create mode 100644 meta/recipes-support/curl/curl/CVE-2023-27535-pre1.patch create mode 100644 meta/recipes-support/curl/curl/CVE-2023-27535_and_CVE-2023-27538.patch create mode 100644 meta/recipes-support/curl/curl/CVE-2023-27536.patch -- 2.34.1 ^ permalink raw reply [flat|nested] 33+ messages in thread
* [OE-core][kirkstone 0/8] Patch review
@ 2022-10-27 2:36 Steve Sakoman
2022-10-28 2:07 ` Tim Orling
0 siblings, 1 reply; 33+ messages in thread
From: Steve Sakoman @ 2022-10-27 2:36 UTC (permalink / raw)
To: openembedded-core
Please review this set of patches for kirkstone and have comments back by
end of day Friday.
This should be the final set of patches for the 4.0.5 release.
Passed a-full on autobuilder:
https://autobuilder.yoctoproject.org/typhoon/#/builders/83/builds/4380
The following changes since commit 4781fee6aea9512b7cb390b76e6f9f0a86a5bd11:
lttng-modules: Fix crash on powerpc64 (2022-10-17 04:30:43 -1000)
are available in the Git repository at:
https://git.openembedded.org/openembedded-core-contrib stable/kirkstone-nut
http://cgit.openembedded.org/openembedded-core-contrib/log/?h=stable/kirkstone-nut
Bruce Ashfield (2):
linux-yocto/5.10: update to v5.10.147
linux-yocto/5.10: update to v5.10.149
Steve Sakoman (1):
Revert "lttng-tools: Upgrade 2.13.4 -> 2.13.8"
Tim Orling (1):
git: upgrade 2.35.4 -> 2.35.5
Vyacheslav Yurkov (2):
files: overlayfs-etc: refactor preinit template
classes: files: Extend overlayfs-etc class
Yash Shinde (2):
binutils: stable 2.38 branch updates
glibc: stable 2.35 branch updates.
meta/classes/overlayfs-etc.bbclass | 5 +++-
meta/files/overlayfs-etc-preinit.sh.in | 23 ++++++++++++++----
meta/recipes-core/glibc/glibc-version.inc | 2 +-
.../binutils/binutils-2.38.inc | 2 +-
.../git/{git_2.35.4.bb => git_2.35.5.bb} | 2 +-
.../linux/linux-yocto-rt_5.10.bb | 6 ++---
.../linux/linux-yocto-tiny_5.10.bb | 8 +++----
meta/recipes-kernel/linux/linux-yocto_5.10.bb | 24 +++++++++----------
...-tools_2.13.8.bb => lttng-tools_2.13.4.bb} | 2 +-
9 files changed, 45 insertions(+), 29 deletions(-)
rename meta/recipes-devtools/git/{git_2.35.4.bb => git_2.35.5.bb} (98%)
rename meta/recipes-kernel/lttng/{lttng-tools_2.13.8.bb => lttng-tools_2.13.4.bb} (98%)
--
2.25.1
^ permalink raw reply [flat|nested] 33+ messages in thread* Re: [OE-core][kirkstone 0/8] Patch review 2022-10-27 2:36 Steve Sakoman @ 2022-10-28 2:07 ` Tim Orling 0 siblings, 0 replies; 33+ messages in thread From: Tim Orling @ 2022-10-28 2:07 UTC (permalink / raw) To: Steve Sakoman; +Cc: openembedded-core [-- Attachment #1: Type: text/plain, Size: 2531 bytes --] On Wed, Oct 26, 2022 at 7:36 PM Steve Sakoman <steve@sakoman.com> wrote: > Please review this set of patches for kirkstone and have comments back by > end of day Friday. > > This should be the final set of patches for the 4.0.5 release. I am intentionally holding off on a Python 3.10.8 upgrade just to let this release get out in a stable manner. > > Passed a-full on autobuilder: > > https://autobuilder.yoctoproject.org/typhoon/#/builders/83/builds/4380 > > The following changes since commit > 4781fee6aea9512b7cb390b76e6f9f0a86a5bd11: > > lttng-modules: Fix crash on powerpc64 (2022-10-17 04:30:43 -1000) > > are available in the Git repository at: > > https://git.openembedded.org/openembedded-core-contrib > stable/kirkstone-nut > > http://cgit.openembedded.org/openembedded-core-contrib/log/?h=stable/kirkstone-nut > > Bruce Ashfield (2): > linux-yocto/5.10: update to v5.10.147 > linux-yocto/5.10: update to v5.10.149 > > Steve Sakoman (1): > Revert "lttng-tools: Upgrade 2.13.4 -> 2.13.8" > > Tim Orling (1): > git: upgrade 2.35.4 -> 2.35.5 > > Vyacheslav Yurkov (2): > files: overlayfs-etc: refactor preinit template > classes: files: Extend overlayfs-etc class > > Yash Shinde (2): > binutils: stable 2.38 branch updates > glibc: stable 2.35 branch updates. > > meta/classes/overlayfs-etc.bbclass | 5 +++- > meta/files/overlayfs-etc-preinit.sh.in | 23 ++++++++++++++---- > meta/recipes-core/glibc/glibc-version.inc | 2 +- > .../binutils/binutils-2.38.inc | 2 +- > .../git/{git_2.35.4.bb => git_2.35.5.bb} | 2 +- > .../linux/linux-yocto-rt_5.10.bb | 6 ++--- > .../linux/linux-yocto-tiny_5.10.bb | 8 +++---- > meta/recipes-kernel/linux/linux-yocto_5.10.bb | 24 +++++++++---------- > ...-tools_2.13.8.bb => lttng-tools_2.13.4.bb} | 2 +- > 9 files changed, 45 insertions(+), 29 deletions(-) > rename meta/recipes-devtools/git/{git_2.35.4.bb => git_2.35.5.bb} (98%) > rename meta/recipes-kernel/lttng/{lttng-tools_2.13.8.bb => > lttng-tools_2.13.4.bb} (98%) > > -- > 2.25.1 > > > -=-=-=-=-=-=-=-=-=-=-=- > Links: You receive all messages sent to this group. > View/Reply Online (#172179): > https://lists.openembedded.org/g/openembedded-core/message/172179 > Mute This Topic: https://lists.openembedded.org/mt/94596838/924729 > Group Owner: openembedded-core+owner@lists.openembedded.org > Unsubscribe: https://lists.openembedded.org/g/openembedded-core/unsub [ > ticotimo@gmail.com] > -=-=-=-=-=-=-=-=-=-=-=- > > [-- Attachment #2: Type: text/html, Size: 5084 bytes --] ^ permalink raw reply [flat|nested] 33+ messages in thread
* [OE-core][kirkstone 0/8] Patch review
@ 2022-08-09 21:27 Steve Sakoman
0 siblings, 0 replies; 33+ messages in thread
From: Steve Sakoman @ 2022-08-09 21:27 UTC (permalink / raw)
To: openembedded-core
Please review this set of patches for the kirkstone 4.0.3 release.
The following changes since commit c33eb7fb1d1e91a005b22b65d221d4b899ec69dc:
openssh: Add openssh-sftp-server to openssh RDEPENDS (2022-08-02 12:32:44 -1000)
are available in the Git repository at:
git://git.openembedded.org/openembedded-core-contrib stable/kirkstone-nut
http://cgit.openembedded.org/openembedded-core-contrib/log/?h=stable/kirkstone-nut
Alexander Kanavin (1):
lttng-modules: update 2.13.3 -> 2.13.4
Bruce Ashfield (5):
linux-yocto/5.10: update to v5.10.135
linux-yocto/5.15: update to v5.15.58
linux-yocto-rt/5.15: update to -rt48 (and fix -stable merge)
linux-yocto/5.15: update to v5.15.59
linux-yocto/5.15: fix reproducibility issues
He Zhe (1):
lttng-modules: Fix build failure for kernel v5.15.58
Sundeep KOKKONDA (1):
glibc : stable 2.35 branch updates
meta/recipes-core/glibc/glibc-version.inc | 2 +-
.../linux/linux-yocto-rt_5.10.bb | 6 +-
.../linux/linux-yocto-rt_5.15.bb | 6 +-
.../linux/linux-yocto-tiny_5.10.bb | 8 +-
.../linux/linux-yocto-tiny_5.15.bb | 6 +-
meta/recipes-kernel/linux/linux-yocto_5.10.bb | 24 +--
meta/recipes-kernel/linux/linux-yocto_5.15.bb | 26 +--
...x-compaction-migratepages-event-name.patch | 37 ----
...oduce-kfree_skb_reason-v5.15.58.v5.1.patch | 53 +++++
...emove-unused-tracepoints-v5.10-v5.15.patch | 44 -----
...g-Append-prev_state-to-tp-args-inste.patch | 59 ------
...vent-allow-same-provider-and-event-n.patch | 48 -----
...g-Don-t-re-read-p-state-when-emittin.patch | 183 ------------------
.../0004-fix-block-remove-genhd.h-v5.18.patch | 45 -----
...emove-REQ_OP_WRITE_SAME-support-v5.1.patch | 79 --------
...ndom-remove-unused-tracepoints-v5.18.patch | 47 -----
...rethook-for-kretprobe-if-possible-v5.patch | 72 -------
...ore-Remove-scsi-scsi_request.h-v5.18.patch | 44 -----
...n-cleanup-the-compaction-trace-event.patch | 106 ----------
...ules_2.13.3.bb => lttng-modules_2.13.4.bb} | 16 +-
20 files changed, 95 insertions(+), 816 deletions(-)
delete mode 100644 meta/recipes-kernel/lttng/lttng-modules/0001-Fix-compaction-migratepages-event-name.patch
create mode 100644 meta/recipes-kernel/lttng/lttng-modules/0001-fix-net-skb-introduce-kfree_skb_reason-v5.15.58.v5.1.patch
delete mode 100644 meta/recipes-kernel/lttng/lttng-modules/0001-fix-random-remove-unused-tracepoints-v5.10-v5.15.patch
delete mode 100644 meta/recipes-kernel/lttng/lttng-modules/0001-fix-sched-tracing-Append-prev_state-to-tp-args-inste.patch
delete mode 100644 meta/recipes-kernel/lttng/lttng-modules/0002-Fix-tracepoint-event-allow-same-provider-and-event-n.patch
delete mode 100644 meta/recipes-kernel/lttng/lttng-modules/0003-fix-sched-tracing-Don-t-re-read-p-state-when-emittin.patch
delete mode 100644 meta/recipes-kernel/lttng/lttng-modules/0004-fix-block-remove-genhd.h-v5.18.patch
delete mode 100644 meta/recipes-kernel/lttng/lttng-modules/0005-fix-scsi-block-Remove-REQ_OP_WRITE_SAME-support-v5.1.patch
delete mode 100644 meta/recipes-kernel/lttng/lttng-modules/0006-fix-random-remove-unused-tracepoints-v5.18.patch
delete mode 100644 meta/recipes-kernel/lttng/lttng-modules/0007-fix-kprobes-Use-rethook-for-kretprobe-if-possible-v5.patch
delete mode 100644 meta/recipes-kernel/lttng/lttng-modules/0008-fix-scsi-core-Remove-scsi-scsi_request.h-v5.18.patch
delete mode 100644 meta/recipes-kernel/lttng/lttng-modules/0010-fix-mm-compaction-cleanup-the-compaction-trace-event.patch
rename meta/recipes-kernel/lttng/{lttng-modules_2.13.3.bb => lttng-modules_2.13.4.bb} (60%)
--
2.25.1
^ permalink raw reply [flat|nested] 33+ messages in threadend of thread, other threads:[~2025-11-03 20:59 UTC | newest] Thread overview: 33+ messages (download: mbox.gz follow: Atom feed -- links below jump to the message on this page -- 2024-02-12 13:54 [OE-core][kirkstone 0/8] Patch review Steve Sakoman 2024-02-12 13:54 ` [OE-core][kirkstone 1/8] libxml2: Fix for CVE-2024-25062 Steve Sakoman 2024-02-12 13:54 ` [OE-core][kirkstone 2/8] go: Fix CVE-2023-45285 and CVE-2023-45287 Steve Sakoman 2024-02-12 13:54 ` [OE-core][kirkstone 3/8] curl: Fix CVE-2023-46219 Steve Sakoman 2024-02-12 13:54 ` [OE-core][kirkstone 4/8] libgit2: Fix CVE-2024-24575 and CVE-2024-24577 Steve Sakoman 2024-02-12 13:54 ` [OE-core][kirkstone 5/8] python3-pycryptodome: Fix CVE-2023-52323 Steve Sakoman 2024-02-12 13:54 ` [OE-core][kirkstone 6/8] systemd: Only add myhostname to nsswitch.conf if in PACKAGECONFIG Steve Sakoman 2024-02-12 13:54 ` [OE-core][kirkstone 7/8] kernel: fix localversion in v6.3+ Steve Sakoman 2024-02-16 1:20 ` Ryan Eatmon 2024-02-16 1:43 ` Steve Sakoman 2024-02-16 1:45 ` Ryan Eatmon 2024-02-16 8:11 ` Andreas Helbech Kleist 2024-02-16 13:51 ` Andreas Helbech Kleist 2024-02-21 14:40 ` Ryan Eatmon 2024-02-22 10:39 ` Andreas Helbech Kleist 2024-02-22 18:38 ` Ryan Eatmon 2024-02-21 14:39 ` Ryan Eatmon 2024-02-12 13:54 ` [OE-core][kirkstone 8/8] ghostscript: correct LICENSE with AGPLv3 Steve Sakoman -- strict thread matches above, loose matches on Subject: below -- 2025-11-03 20:59 [OE-core][kirkstone 0/8] Patch review Steve Sakoman 2025-10-17 20:43 Steve Sakoman 2025-08-13 21:28 Steve Sakoman 2025-04-01 22:36 Steve Sakoman 2024-08-04 17:08 Steve Sakoman 2024-05-24 12:14 Steve Sakoman 2024-03-12 13:53 Steve Sakoman 2023-11-29 23:04 Steve Sakoman 2023-09-13 14:30 Steve Sakoman 2023-08-27 20:52 Steve Sakoman 2023-07-24 2:33 Steve Sakoman 2023-04-22 15:54 Steve Sakoman 2022-10-27 2:36 Steve Sakoman 2022-10-28 2:07 ` Tim Orling 2022-08-09 21:27 Steve Sakoman
This is a public inbox, see mirroring instructions for how to clone and mirror all data and code used for this inbox