Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Fix OpenSSL::PKey.read that cannot parse PKey in the FIPS mode. #615

Merged
merged 3 commits into from
Jun 1, 2023

Conversation

junaruga
Copy link
Member

@junaruga junaruga commented Apr 17, 2023

This PR is to fix the following issue that is managed as one of the issues at #603 on the FIPS mode. This PR is based on the PR #608. So, please consider reviewing and merging the #608 first. Then I can rebase this PR on the latest master branch.

$ openssl genrsa -out key.pem 4096

$ ruby -e "require 'openssl'; OpenSSL::PKey.read(File.read('key.pem'))"
-e:1:in `read': Could not parse PKey (OpenSSL::PKey::PKeyError)
  from -e:1:in `<main>'

There are some commits on this PR. The first 3 commits are basically from the PR #603. Then the 4th commit is the main commit to fix this issue. The patch is the same with what I mentioned at #603 (comment). The 4th and 5th commits are to run the unit test test/openssl/test_pkey.rb.

https://github.com/junaruga/openssl/actions/runs/4681990779/jobs/8295272065#step:11:732

Before the main 4rd commit, the test_generic_oid_inspect and test_to_text was failing in the test/openssl/test_pkey.rb. So, running test/openssl/test_pkey.rb can test that this issue was fixed.

Note that the 3 tests test_ed25519, test_x25519 and test_compare? in the test/openssl/test_pkey.rb are still failing due to other issues. And I am working on it at the #603 (comment).

I confirmed that CI passed on my forked repository.
https://github.com/junaruga/openssl/actions/runs/4724029713/jobs/8380737782#step:12:64

@junaruga
Copy link
Member Author

Rebased on the latest master branch.

@junaruga
Copy link
Member Author

Tomorrow I manually will check if the new compiler warnings are printed or not on each case in the CI.

… mode.

This commit is a workaround to avoid the error below that the
`OpenSSL::PKey.read` fails with the OpenSSL 3.0 FIPS mode.

```
$ openssl genrsa -out key.pem 4096

$ ruby -e "require 'openssl'; OpenSSL::PKey.read(File.read('key.pem'))"
-e:1:in `read': Could not parse PKey (OpenSSL::PKey::PKeyError)
  from -e:1:in `<main>'
```

The root cause is on the OpenSSL side. The `OSSL_DECODER_CTX_set_selection`
doesn't apply the selection value properly if there are multiple providers, and
a provider (e.g.  "base" provider) handles the decoder implementation, and
another provider (e.g. "fips" provider) handles the keys.

The workaround is to create `OSSL_DECODER_CTX` variable each time without using
the `OSSL_DECODER_CTX_set_selection`.
We want to run the unit tests in the FIPS mode too.
It's to test the `OpenSSL::PKey.read` in the `test/openssl/test_pkey.rb`.

I added the pending status to the following tests failing on the FIPS mode
case in the `test/openssl/test_pkey.rb`.

* `test_ed25519`
* `test_x25519`
* `test_compare?`
@junaruga
Copy link
Member Author

junaruga commented Jun 1, 2023

Tomorrow I manually will check if the new compiler warnings are printed or not on each case in the CI.

Manually checking the compiler warnings on the CI is very hard. First, I identified what CI cases I need to check on this PR's CI. To find the cases, I executed this PR and master branch's CIs with one commit #631 to check the compiler warnings.

diff --git a/.github/workflows/test.yml b/.github/workflows/test.yml
index becf990..e553d06 100644
--- a/.github/workflows/test.yml
+++ b/.github/workflows/test.yml
@@ -19,6 +19,7 @@ jobs:
         # ubuntu-22.04 uses OpenSSL 3.0, ubuntu-20.04 uses OpenSSL 1.1.1
         os: [ ubuntu-22.04, ubuntu-20.04, macos-latest, windows-latest ]
         ruby: ${{ fromJson(needs.ruby-versions.outputs.versions) }}
+        fail_by_warn: [ true ]
         exclude:
           # uses non-standard MSYS2 OpenSSL 3 package
           - { os: windows-latest, ruby: head }
@@ -47,6 +48,10 @@ jobs:
         run:  echo "MAKEFLAGS=V=1" >> $GITHUB_ENV
         if: runner.os == 'Linux' || runner.os == 'macOS'
 
+      - name: set flags to fail by the compiler warnings.
+        run:  echo "RUBY_OPENSSL_EXTCFLAGS=-Werror" >> $GITHUB_ENV
+        if: matrix.fail_by_warn
+
       - name: compile
         run:  rake compile -- --enable-debug
 
@@ -75,6 +80,7 @@ jobs:
           - libressl-3.5.3
           - libressl-3.6.1
           - libressl-3.7.0 # Development release
+        fail_by_warn: [ true ]
         fips_enabled: [ false ]
         include:
           - { os: ubuntu-latest, ruby: "3.0", openssl: openssl-3.0.8, fips_enabled: true, append_configure: 'enable-fips', name_extra: 'fips' }
@@ -138,6 +144,10 @@ jobs:
         run:  echo "MAKEFLAGS=V=1" >> $GITHUB_ENV
         if: runner.os == 'Linux' || runner.os == 'macOS'
 
+      - name: set flags to fail by the compiler warnings.
+        run:  echo "RUBY_OPENSSL_EXTCFLAGS=-Werror" >> $GITHUB_ENV
+        if: matrix.fail_by_warn
+
       - name: compile
         run:  rake compile -- --enable-debug --with-openssl-dir=$HOME/.openssl/${{ matrix.openssl }}

The CI result for this PR + a commit of checking the compiler warnings
https://github.com/junaruga/openssl/actions/runs/5144488890

The CI result for the master branch (8aee87305d6ef7b353e682e5dff9180b047948bf) + a commit of checking the compiler warnings
https://github.com/ruby/openssl/actions/runs/5142916181?pr=631

  • ubuntu-22.04 2.6
  • ubuntu-22.04 3.1
  • ubuntu-22.04 3.2
  • ubuntu-22.04 head
  • ubuntu-22.04 truffleruby-head
  • ubuntu-20.04 2.6
  • ubuntu-20.04 truffleruby-head
  • macos-latest 2.6
  • macos-latest truffleruby-head

Then I will be checking if there are the compiler warnings in the cases in this PR's CI.

@junaruga
Copy link
Member Author

junaruga commented Jun 1, 2023

Then I will be checking if there are the compiler warnings in the cases in this PR's CI.

I checked it manually. And I confirmed there are no new compiler warnings by this PR.

  • ubuntu-22.04 2.6: There are warnings. But these are not new by this PR.
     ../../../../ext/openssl/ossl_ssl.c: In function ‘ossl_start_ssl’:
    ../../../../ext/openssl/ossl_ssl.c:1761:5: warning: ISO C90 forbids mixed declarations and code [-Wdeclaration-after-statement]
     1761 |     VALUE io = rb_attr_get(self, id_i_io);
          |     ^~~~~
    ../../../../ext/openssl/ossl_ssl.c: In function ‘ossl_ssl_read_internal’:
    ../../../../ext/openssl/ossl_ssl.c:1948:5: warning: ISO C90 forbids mixed declarations and code [-Wdeclaration-after-statement]
     1948 |     VALUE io = rb_attr_get(self, id_i_io);
          |     ^~~~~
    ../../../../ext/openssl/ossl_ssl.c: In function ‘ossl_ssl_write_internal’:
    ../../../../ext/openssl/ossl_ssl.c:2049:5: warning: ISO C90 forbids mixed declarations and code [-Wdeclaration-after-statement]
     2049 |     VALUE io = rb_attr_get(self, id_i_io);
          |     ^~~~~
    
  • ubuntu-22.04 3.1: No warnings.
  • ubuntu-22.04 3.2: No warnings.
  • ubuntu-22.04 head: No warnings.
  • ubuntu-22.04 truffleruby-head: There is a warning. But it is not new by this PR.
    ../../../../ext/openssl/ossl_pkey.c:342:22: warning: implicit truncation from 'int' to a one-bit wide bit-field changes value from 1 to -1 [-Wsingle-bit-bitfield-constant-conversion]
        arg->interrupted = 1;
                         ^ ~
    1 warning generated.
    
  • ubuntu-20.04 2.6: There are warnings. But these are not new by this PR.
    ../../../../ext/openssl/ossl_ssl.c: In function ‘ossl_start_ssl’:
    ../../../../ext/openssl/ossl_ssl.c:1761:5: warning: ISO C90 forbids mixed declarations and code [-Wdeclaration-after-statement]
     1761 |     VALUE io = rb_attr_get(self, id_i_io);
          |     ^~~~~
    ../../../../ext/openssl/ossl_ssl.c: In function ‘ossl_ssl_read_internal’:
    ../../../../ext/openssl/ossl_ssl.c:1948:5: warning: ISO C90 forbids mixed declarations and code [-Wdeclaration-after-statement]
     1948 |     VALUE io = rb_attr_get(self, id_i_io);
          |     ^~~~~
    ../../../../ext/openssl/ossl_ssl.c: In function ‘ossl_ssl_write_internal’:
    ../../../../ext/openssl/ossl_ssl.c:2049:5: warning: ISO C90 forbids mixed declarations and code [-Wdeclaration-after-statement]
     2049 |     VALUE io = rb_attr_get(self, id_i_io);
          |     ^~~~~
    
  • ubuntu-20.04 truffleruby-head: There is a warning. But it is not new by this PR.
    ../../../../ext/openssl/ossl_pkey.c:342:22: warning: implicit truncation from 'int' to a one-bit wide bit-field changes value from 1 to -1 [-Wsingle-bit-bitfield-constant-conversion]
        arg->interrupted = 1;
                         ^ ~
    
  • macos-latest 2.6: There are many warnings and notes for several .c files. And checking only the warnings for openssl/ossl_pkey.c that is modified in this PR, the links is https://github.com/ruby/openssl/actions/runs/5143196869/jobs/9257919215?pr=615#step:6:1697 and there are 4 warnings by the -Wcompound-token-split-by-macro below. The result is same with the current master branch's CI result https://github.com/ruby/openssl/actions/runs/5145354568/jobs/9262893433#step:6:1699.
    clang -I. -I/Users/runner/hostedtoolcache/Ruby/2.6.10/x64/include/ruby-2.6.0/x86_64-darwin19 -I/Users/runner/hostedtoolcache/Ruby/2.6.10/x64/include/ruby-2.6.0/ruby/backward -I/Users/runner/hostedtoolcache/Ruby/2.6.10/x64/include/ruby-2.6.0 -I../../../../ext/openssl -DRUBY_EXTCONF_H=\"extconf.h\" -I/Users/runner/hostedtoolcache/Ruby/2.6.10/x64/openssl/include -I/Users/runner/hostedtoolcache/Ruby/2.6.10/x64/include -DENABLE_PATH_CHECK=0 -D_XOPEN_SOURCE -D_DARWIN_C_SOURCE -D_DARWIN_UNLIMITED_SELECT -D_REENTRANT   -fno-common -O3 -ggdb3 -Wall -Wextra -Wdeclaration-after-statement -Wdeprecated-declarations -Wdivision-by-zero -Wimplicit-function-declaration -Wimplicit-int -Wmisleading-indentation -Wpointer-arith -Wshorten-64-to-32 -Wwrite-strings -Wmissing-noreturn -Wno-constant-logical-operand -Wno-long-long -Wno-missing-field-initializers -Wno-overlength-strings -Wno-parentheses-equality -Wno-self-assign -Wno-tautological-compare -Wno-unused-parameter -Wno-unused-value -Wunused-variable -Wextra-tokens  -fno-common -pipe  -o ossl_pkey.o -c ../../../../ext/openssl/ossl_pkey.c
    ../../../../ext/openssl/ossl_pkey.c:260:28: warning: '(' and '{' tokens introducing statement expression appear in different macro expansion contexts [-Wcompound-token-split-by-macro]
        rb_block_call(args[1], rb_intern("each"), 0, NULL,
                               ^~~~~~~~~~~~~~~~~
    ...<snip>...
    ../../../../ext/openssl/ossl_pkey.c:260:28: warning: '}' and ')' tokens terminating statement expression appear in different macro expansion contexts [-Wcompound-token-split-by-macro]
        rb_block_call(args[1], rb_intern("each"), 0, NULL,
                               ^~~~~~~~~~~~~~~~~
    ...<snip>...
    ../../../../ext/openssl/ossl_pkey.c:561:9: warning: '(' and '{' tokens introducing statement expression appear in different macro expansion contexts [-Wcompound-token-split-by-macro]
        if (OSSL_PKEY_IS_PRIVATE(obj))
            ^~~~~~~~~~~~~~~~~~~~~~~~~
    ...<snip>...
    ../../../../ext/openssl/ossl_pkey.c:561:9: warning: '}' and ')' tokens terminating statement expression appear in different macro expansion contexts [-Wcompound-token-split-by-macro]
        if (OSSL_PKEY_IS_PRIVATE(obj))
            ^~~~~~~~~~~~~~~~~~~~~~~~~
    ...<snip>...
    4 warnings generated.
    
  • macos-latest truffleruby-head: There is a warning. But it is not new by this PR.
    ../../../../ext/openssl/ossl_pkey.c:342:22: warning: implicit truncation from 'int' to a one-bit wide bit-field changes value from 1 to -1 [-Wsingle-bit-bitfield-constant-conversion]
        arg->interrupted = 1;
                         ^ ~
    

@junaruga junaruga merged commit 7e411b4 into ruby:master Jun 1, 2023
42 checks passed
@junaruga junaruga deleted the wip/fips-read branch June 1, 2023 14:54
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
None yet
Development

Successfully merging this pull request may close these issues.

None yet

1 participant