| Age | Commit message (Collapse) | Author | Files | Lines |
|---|
|
RespondToUnauthenticatedClients allows users to explicitly select mTLS
as their only authentication mechanism, thus significantly reducing
their code exposure to unauthenticated clients.
From the Redfish specification
```
The RespondToUnauthenticatedClients property within the
ClientCertificate property within the MFA property of the AccountService
resource controls the response behavior when an invalid certificate is
provided by the client.
• If the property contains true or is not
supported by the service, the service shall not fail the TLS handshake.
This is to allow the service to send error messages or unauthenticated
resources to the client.
• If the property contains false , the service
shall fail the TLS handshake.
```
This commit implements that behavior.
This also has some added benefits in that we no longer have to check the
filesystem for every connection, as TLS is controlled explicitly, and
not whether or not a root cert is in place.
Note, this also implements a TODO to disable cookie auth when using
mTLS. Clients can still use IsAuthenticated to determine if they are
authenticated on request.
Tested:
Run scripts/generate_auth_certs.py to set up a root certificate and
client certificate. This verifies that mTLS as optional has not been
broken. Script succeeds.
```
PATCH /redfish/v1/AccountService
{"MultiFactorAuth": {"ClientCertificate": {"RespondToUnauthenticatedClients": false}}}
```
GET /redfish/v1
without a client certificate now fails with an ssl verification error
GET /redfish/v1
with a client certificate returns the result
```
PATCH /redfish/v1/AccountService
{"MultiFactorAuth": {"ClientCertificate": {"RespondToUnauthenticatedClients": false}}}
With certificate returns non mTLS functionality.
```
Change-Id: I5a9d6d6b1698bff83ab62b1f760afed6555849c9
Signed-off-by: Ed Tanous <ed@tanous.net>
|
|
Because of recent changes to how dbus authentication is done, Requests
might be moved out before they can be used. This commit is an attempt
to mitigate the problem without needing to revert that patch.
This commit does two relatively distinct things.
First, it moves basic auth types to a model where they're timed out
instead of removed on destruction. This removes the need for a Request
object to track that state, and arguably gives better behavior, as
basic auth sessions will survive through the timeout.
To prevent lots of basic auth sessions getting created, a basic auth
session is reused if it was:
1. Created by basic auth previously.
2. Created by the same user.
3. Created from the same source IP address.
Second, both connection classes now store the accept, and origin headers
from the request in the connection class itself, removing the need for
them.
Tested: HTML page now loads when pointing at a redfish URL with a
browser.
Change-Id: I623b43cbcbb43d9e65b408853660be09a5edb2b3
Signed-off-by: Ed Tanous <ed@tanous.net>
|
|
Redfish AccountService[1] defines methods for selecting how to map a
certificate CommonName attribute to a user. These are intended to be a
patch parameter.
This commit implements the Redfish defined schemas; The parsing mode is
stored in the bmcweb persistent configuration file as an integer enum,
with Mapping to the Redfish schema.
To handle OEM specific parsing modes, an enum value of 100+ is defined
to allow the additional OEM parameters. Unfortunately, Redfish doesn't
have a way to represent these today, so those modes are currently not
selectable at runtime.
Now that things are runtime selectable, this obsoletes the option
mutual-tls-common-name-parsing, as it is not longer required at compile
time.
Tested:
GET /redfish/v1/AccountService
returns MultiFactorAuth/ClientCertificate/CertificateMappingAttribute
PATCH /redfish/v1/AccountService
```
{"MultiFactorAuth": {"ClientCertificate": {"CertificateMappingAttribute":"CommonName"}}}
```
Returns 200
[1] https://github.com/DMTF/Redfish-Publications/blob/5b217908b5378b24e4f390c063427d7a707cd308/csdl/AccountService_v1.xml#L1631
Change-Id: I67db0dfa5245a9da973320aab666d12dbd9229e4
Signed-off-by: Ed Tanous <ed@tanous.net>
|
|
Using Redfish-standard X-Auth-Token authentication is less secure
(against injected JS code) compared to an HttpOnly (not available to the
JS VM) SESSION cookie. Currently webui-vue authenticates connections to
WebSocket URIs not only by a JS-accessible token (passed as subprotocol
when upgrading to WS) but also via a SESSION cookie (even though it is
not subject to CORS policy).
To allow WebSocket-based functionality (IP KVM, SOL, VM) after creating
a Session object send a set of cookies instead of the X-Auth-Token
header if the request was made by webui-vue (detected by presence of
"X-Requested-With" header).
Factor out cookie setting and clearing functions and use explicit Path=/
attribute as the cookies are valid for the whole server, not just the
path of the endpoint they were created by.
Not specifying Path was functional for /login endpoint because
https://www.rfc-editor.org/rfc/rfc6265#section-5.3 point 7 for this case
says "set the cookie's path to the default-path of the request-uri" and
https://www.rfc-editor.org/rfc/rfc6265#section-5.1.4 tells how to
compute the default path. Basically, it was a "happy coincidence" that
/login defaults to / for the Path, if it was /openbmc/login then the
cookies would have been set to Path=/openbmc and not work at all for
/redfish/v1 endpoints.
Tested: Redfish-Service-Validator doesn't see a difference. Runtime
testing logging in via Sessions endpoint, getting data, using websockets
and logging out against webui-vue with a corresponding change while
carefully observing Request and Response headers. Creating a session
with curl without the special header shows just X-Auth-Token and no
cookies in the response.
Change-Id: I0b1774e586671874bb79f115e9cddf194f9ea653
Signed-off-by: Paul Fertser <fercerpav@gmail.com>
|
|
This file has several long lambdas, and some functions that should be
broken down into smaller components. This is a refactor that should
make no functional changes.
Tested: Webui-vue loads correctly, shows static files are working as
designed.
Change-Id: Ie272a15fd935823743303c287a475114a734feb5
Signed-off-by: Ed Tanous <ed@tanous.net>
|
|
When the session is just getting created the normal privileges
validation workflow isn't executed and so the current role remains
unknown. Fix this by refactoring dbus_privileges.hpp to allow obtaining
the information from phosphor-user-manager late in the request
processing.
Tested: Redfish Service Validator passes.
Creating a session for local user:
```
$ curl -k -H "Content-Type: application/json" -X POST https://172.41.1.250:18080/redfish/v1/SessionService/Sessions -d '{"UserName":"root", "Password":"0penBmc"}'
{
"@odata.id": "/redfish/v1/SessionService/Sessions/lfFsCNjshV",
"@odata.type": "#Session.v1_7_0.Session",
"ClientOriginIPAddress": "172.40.1.4",
"Description": "Manager User Session",
"Id": "lfFsCNjshV",
"Name": "User Session",
"Roles": [
"Administrator"
],
"UserName": "root"
}
```
Creating a session for remote user mapped to Operator:
```
$ curl -k -H "Content-Type: application/json" -X POST https://172.41.1.250:18080/redfish/v1/SessionService/Sessions -d '{"UserName":ldap_sync", "Password":"ldap_password"}'
{
"@odata.id": "/redfish/v1/SessionService/Sessions/qVffc4ePJK",
"@odata.type": "#Session.v1_7_0.Session",
"ClientOriginIPAddress": "172.40.1.4",
"Description": "Manager User Session",
"Id": "qVffc4ePJK",
"Name": "User Session",
"Roles": [
"Operator"
],
"UserName": "ldap_sync"
}
```
Fixes: https://github.com/openbmc/bmcweb/issues/280
Fixes: ce22f6099e7e28ae26591348bf484ebedbc1ed42
Change-Id: If76c43563244e3819ee3fbc60d9df7f6a21c1fa3
Signed-off-by: Paul Fertser <fercerpav@gmail.com>
|
|
nlohmann::json::items() throws an exception if the object in question is
not a json object. This has the potential to cause problems, and isn't
in line with the standard that we code against.
Replace all uses of items with iterating the nlohmann::json::object_t.
This adds a new error check for pulling the object_t out of the
nlohmann::json object before each iteration, to ensure that we're
handling errors.
Tested: Redfish service validator passes.
Change-Id: I2934c9450ec296c76544c2a7c5855c9b519eae7f
Signed-off-by: Ed Tanous <ed@tanous.net>
|
|
WebUI cannot login cause by 38221509e6cc06a6897e8b9b2aa049a70c033840.
The root URL ("/") represents the top-level directory of a website.
Removing the trailing slash from the root URL could lead to incorrect
representations.
Therefore, add to check that the URL is not the root before removing the
trailing slash.
Fixes #279
Tested: Use Chrome to access
https://${BMC_IP}.
Success login WebUI.
Change-Id: I19527ee785f550ba5aeafe6e94e8521ab508817c
Signed-off-by: VinceChang6637 <vince_chang@aspeedtech.com>
|
|
A bug was found in CI where the server.pem was a directory vs. the
expected file.
Tested:
- None, simple bug
Change-Id: Ib0abafe34864f505e7c677457bae89e5aefca394
Signed-off-by: Andrew Geissler <geissonator@yahoo.com>
|
|
The method of creating a random ID from an openssl random generator of a
particular length is something that is generally useful, and something
we can write unit tests for. Add it.
Tested:
Redfish service validator login flows work correctly in redfish service
validator.
Change-Id: Ic3b58d33f1421f3eb39e2d57585958f87f6fb8ea
Signed-off-by: Ed Tanous <ed@tanous.net>
|
|
http_client currently does not uses mtls client certificates. This is
a good feature for both authentication and authorization purpose.
It will help external servers to trust the identity of bmc for better
security.This patch will add MTLS client certificate support for bmcweb
This is a needed feature to support secure redfish aggregation between
BMCs.
To support secure aggregation BMCs should be provisioned with CA signed
certificate with an authorized username as the subject name field of the
certificate.
With the support of strong MTLS authentication from Bmcweb server we can
use the MTLS path to enable secure redfish aggregation among BMCs. This
can avoid complexities and extra API calls needed for token based
approach.
Tested by:
Aggregation Test1:
1) Setup two instance of romulus qemu session at different ports.This
will act as two BMCs
2) Installed CA root certificates at /etc/ssl/certs/authority in both
BMCs
3) Installed server.pem and client.pem entity certificates signed by the
root CA at /etc/ssl/certs/https folder in both BMCs
4) Enable aggregation for Bmcweb.
5) Fired several redfish queries to BMC1
Result
Observed that the aggregation worked fine. User session created using
username mentined in the CN field of certificate.
Aggregation Test2:
Followed same steps from Aggregation Test1 with modification in step 3
In step3 installed only the server.pem.
Result
Bmcweb ran as usual. But aggregation failed to collect resources from
BMC2. No crash observed.
Redfish Event Test:
Subscribed for redfish events using test server.
Fired redfish test events from BMC.
Result:
Events reached server successfully.
Change-Id: Id8cccf9beec77da0f16adb72d52f3adf46347d06
Signed-off-by: Abhilash Raju <abhilash.kollam@gmail.com>
Signed-off-by: Ed Tanous <etanous@nvidia.com>
|
|
When run from a development PC, we shouldn't REQUIRE that the cert
directory exists or is writable.
This commit reworks the SSL cert generation to generate a string with
the certification info, instead of writing it to disk and reading it
back. This allows bmcweb to start up in read-only environments, or
environments where there isn't access to the key information.
Tested: Launching the application on a dev desktop without an ssl
directory present no longer crashes.
Change-Id: I0d44eb1ce8d298986c5560803ca2d72958d3707c
Signed-off-by: Ed Tanous <ed@tanous.net>
|
|
Metadata payloads are no longer accessible without authentication due to
a regression caused by 090ab8e1042e14f7e5e02572ae2a2102677f1f00.
Add /redfish/v1/$metadata to the allow list, and use this as an
opportunity to refactor the isOnAllowList() code and simplify it.
Fixes #277
Tested: Redfish protocol validator $metadata tests now pass again.
```
curl -vvvv -k https://192.168.7.2/redfish/v1/\$metadata
```
Now succeeds.
Change-Id: I0cc3492f6184b2cad3281b22535d5089709c16a4
Signed-off-by: Ed Tanous <etanous@nvidia.com>
|
|
In the past, we've tried to erradicate the use of
nlohmann::json(initiatlizer_list<...>) because it bloats binary sizes,
as every type is given a new nlohmann constructor.
This commit hunts down the last few places where we call this. There is
still 2 remaining in openbmc_dbus_rest after this, but those are variant
accesses that are difficult to triage, and considering it's a less used
api, they're left as is.
Tested: WIP
Change-Id: Iaac24584bb78bb238da69010b511c1d598bd38bc
Signed-off-by: Ed Tanous <ed@tanous.net>
|
|
In the initial implementation of metadata indexing the bmc knew at
compile time what schemas it could potentially publish. bmcweb took the
approach of adding all schemas of all versions to the $metadata
resource. Since that was made, two major changes have happened.
First, Redfish has added significantly more versions of each schema, as
well as significantly more schemas to the point where the metadata index
is now 213KB. While this file compresses fairly well, the size is
obvious from the large amount of time that redfish service validator
takes to parse the schemas, compared to actually acquiring BMC redfish
resources.
Second, aggregation was added, where an aggregated Redfish service might
implement any number of schemas, including OEM ones.
In an effort to fix this, this patch takes the compile-time algorithm in
update_schemas.py, and moves it into bmcweb itself, parsing the files on
disk as needed on demand. This has some immediate benefits; First, is
that now schemas can be potentially installed from anywhere, not only
from within the bmcweb build, and they will be resolved at runtime.
Second, patches that want to add support for a given schema need to only
symlink the schema into the correct folder, without needing to rerun
update_schemas.py. This saves time in review.
Finally, this opens to door to reducing the schema versions present in
the metadata to the unique set of only what this bmcweb instance, and
its aggregated BMCs expose.
Tested: Redfish service validator passes. Need A/B checking to verify
the file is byte for byte the same.
GET /redfish/v1/$metadata returns what looks like sane results, with a
correct content-type.
Unit tests require the use of TemporaryFileHandle, so that class is
moved into a more general folder, outside of test/http.
Change-Id: I326159099c6b6c4056023b2e173c5f074ed88ce1
Signed-off-by: Ed Tanous <ed@tanous.net>
|
|
When a user's password is changed, existing Redfish sessions for that
user, created with the old password, continue to work.
As per OWASP session management, "The session ID must be renewed or
regenerated by the web application after any privilege level change
within the associated user session... Common scenarios to consider
include; password changes, permission changes, or switching from a
regular user role to an administrator role within the web application."
[1] https://cheatsheetseries.owasp.org/cheatsheets/Session_Management_Cheat_Sheet.html
This commit removes existing user sessions when the user's password is
changed. This commit leaves the current session in place though a new
removeSessionsByUsernameExceptSession().
This commit doesn't completely get us fully to what owasp says but is a
start.
Tested:
Create some users:
```
curl -k -v -X POST -H "Content-Type: application/json" \
https://$bmc/redfish/v1/AccountService/Accounts/ -d \
'{"UserName":"testadminuser","Password":"<password>","RoleId":"Administrator","Enabled":true}'
```
Using basic auth was able to update own password and another user's
password.
Using token auth, verified the current session did not get deleted but
other sessions from that user did.
```
curl -k -H "Content-Type: application/json" -X POST -D headers.txt \
https://${bmc}/redfish/v1/SessionService/Sessions -d \
'{"UserName":"testadminuser", "Password":"<password>"}'
```
```
curl -k -v -X PATCH -H "X-Auth-Token: $token" \
-H "Content-Type:application/json" \
https://$bmc/redfish/v1/AccountService/Accounts/testadminuser \
-d '{"Password":"<password>"}'
```
Verified when changing another user's password all sessions were
dropped.
Change-Id: I4de60b84964a6b29c021dc3a2bece9ed4bc09eac
Signed-off-by: Ravi Teja <raviteja28031990@gmail.com>
|
|
Broken by [1], interface and path were incorrect after changing
getManagedObjects to getAllProperties.
Tested:
Mounting in proxy mode works
[1]: https://github.com/openbmc/bmcweb/commit/e4b32753a7ce7ca436bf751f390ee01f02b9efd5
Change-Id: I06f36e8fc864fe13200d5d13a12639ebba9d9be1
Signed-off-by: Boleslaw Ogonczyk Makowski <boleslawx.ogonczyk-makowski@intel.com>
|
|
Static analysis flags that we're ignoring a return value here. Use it.
Tested: Code compiles.
Change-Id: I2b37286b5a7b549b483ed5669fa0c24a628adc98
Signed-off-by: Ed Tanous <ed@tanous.net>
|
|
CSRF option got inverted. Fix it.
Tested: Code compiles.
Change-Id: Ibfc56ef2ce8d065aa7dad836e3d4a5edc5632926
Signed-off-by: Ed Tanous <ed@tanous.net>
|
|
This is flagging static analysis issues, because we don't reserve a
quadraticly sized increment. This is a micro optimization that never
made much sense.
Tested: Old code. Inspection only.
Change-Id: I442628751558616c24123dba66aab448a4c24039
Signed-off-by: Ed Tanous <ed@tanous.net>
|
|
9f217c26f58c0a99c18e7cac7b095dcf6068562d
had a merge conflict that was resolved incorrectly with
25b54dba775b31021a3a4677eb79e9771bcb97f7
The line returning the cookie session got dropped in that merge
conflict. Restore it.
Tested: Webui can log in, and cookie auth works again.
Ideally in the future we'd have some tests for places like this where
we've gone outside the redfish spec.
Change-Id: I7740e19dac4d0dae5c5c9b27a5b8699a4751fd6f
Signed-off-by: Ed Tanous <ed@tanous.net>
|
|
Change-Id: Ie7729554e438f3af9ddb58941297525c50c6b419
Signed-off-by: Ed Tanous <ed@tanous.net>
|
|
Static analysis finds these two places that we don't check error codes.
Check them.
Tested: Deprecated code. Inspection only.
Change-Id: I92c238c5a4b1f51c5c6855c59f4f943855c4e50f
Signed-off-by: Ed Tanous <ed@tanous.net>
|
|
Args captured by logging functions should be captured by rvalue, and use
std::forward to get perfect forwarding. In addition, separate out the
various std::out lines.
While we're here, also try to optimize a little. We should ideally be
writing each log line to the output once, and ideally not use iostreams,
which induce a lot of overhead.
Similar to spdlog[1] (which at one point this codebase used), construct
the string, then call fwrite and fflush once, rather than calling
std::cout repeatedly.
Now that we don't have a dependency on iostreams anymore, we can remove
it from the places where it has snuck in.
Tested:
Logging still functions as before. Logs present.
[1] https://github.com/gabime/spdlog/blob/27cb4c76708608465c413f6d0e6b8d99a4d84302/include/spdlog/sinks/stdout_sinks-inl.h#L70C7-L70C13
Change-Id: I1dd4739e06eb506d68989a066d122109b71b92cd
Signed-off-by: Ed Tanous <ed@tanous.net>
|
|
Fix some obvious missed headers.
Tested: It builds, no other testing.
Change-Id: I8cfd95e16eca38c75dcf76fc974ed384d13c6757
Signed-off-by: Gunnar Mills <gmills@us.ibm.com>
|
|
This is an attempt to solve a class of use-after-move bugs on the
Request objects which have popped up several times. This more clearly
identifies code which owns the Request objects and has a need to keep it
alive. Currently it's just the `Connection` (or `HTTP2Connection`)
(which needs to access Request headers while sending the response), and
the `validatePrivilege()` function (which needs to temporarily own the
Request while doing an asynchronous D-Bus call). Route handlers are
provided a non-owning `Request&` for immediate use and required to not
hold the `Request&` for future use.
Tested: Redfish validator passes (with a few unrelated fails).
Redfish URLs are sent to a browser as HTML instead of raw JSON.
Change-Id: Id581fda90b6bceddd08a5dc7ff0a04b91e7394bf
Signed-off-by: Jonathan Doman <jonathan.doman@intel.com>
Signed-off-by: Ed Tanous <ed@tanous.net>
|
|
The configuration options that exist in bmcweb are an amalgimation of
CROW options, CMAKE options using #define, pre-bmcweb ifdef mechanisms
and meson options using a config file. This history has led to a lot of
different ways to configure code in the codebase itself, which has led
to problems, and issues in consistency.
ifdef options do no compile time checking of code not within the branch.
This is good when you have optional dependencies, but not great when
you're trying to ensure both options compile.
This commit moves all internal configuration options to:
1. A namespace called bmcweb
2. A naming scheme matching the meson option. hyphens are replaced with
underscores, and the option is uppercased. This consistent transform
allows matching up option keys with their code counterparts, without
naming changes.
3. All options are bool true = enabled, and any options with _ENABLED or
_DISABLED postfixes have those postfixes removed. (note, there are
still some options with disable in the name, those are left as-is)
4. All options are now constexpr booleans, without an explicit compare.
To accomplish this, unfortunately an option list in config/meson.build
is required, given that meson doesn't provide a way to dump all options,
as is a manual entry in bmcweb_config.h.in, in addition to the
meson_options. This obsoletes the map in the main meson.build, which
helps some of the complexity.
Now that we've done this, we have some rules that will be documented.
1. Runtime behavior changes should be added as a constexpr bool to
bmcweb_config.h
2. Options that require optionally pulling in a dependency shall use an
ifdef, defined in the primary meson.build. (note, there are no
options that currently meet this class, but it's included for
completeness.)
Note, that this consolidation means that at configure time, all options
are printed. This is a good thing and allows direct comparison of
configs in log files.
Tested: Code compiles
Server boots, and shows options configured in the default build. (HTTPS,
log level, etc)
Change-Id: I94e79a56bcdc01755036e4e7278c7e69e25809ce
Signed-off-by: Ed Tanous <ed@tanous.net>
|
|
nbd proxy and vm websocket options got reversed in
36c0f2a35e670a4b798b7b42fd18455085e9d9c0
Change them back.
Change-Id: I7c54e66f88aee956bd20f2139d110e64998a4ef5
Signed-off-by: Ed Tanous <ed@tanous.net>
|
|
Tested: Need help
Change-Id: I28cc1626212ec746b5345490ec285706eb386e65
Signed-off-by: Ed Tanous <ed@tanous.net>
|
|
As much as the two vm implementations SEEM different, the differences
largely lie in how we're getting the nbd proxy socket. One is relying
on launching a process (nbd-proxy), the other is getting the fd from
dbus. Given [1] exists and is in process, we need to have a plan for
getting these two VM implementations into one, once that patchset is
complete.
This commit: Splits the vm-websocket option into vm-websocket-provider,
providing two options, nbd-proxy, and virtual-media (the names of the
respective apps). To accomplish this, it moves the contents of
nbd-proxy into include/vm-websocket, so we can compare the similarities
and start consolidating.
The longer term intent is that the nbd-proxy option will be completely
removed, and the code deleted. This has the additional advantage that
we will no longer require the boost::process dependency, as all info
will be available on dbus.
As part of this, the nbd proxy websocket is also registered at /vm/0/0,
to be backward compatible with the old interfaces.
Tested: Code compiles. Need some help here.
[1] https://gerrit.openbmc.org/c/openbmc/jsnbd/+/49944
Change-Id: Iedbca169ea40d45a8775f843792b874a248bb594
Signed-off-by: Ed Tanous <ed@tanous.net>
|
|
In the change made to move to std::format, we defined some custom type
formatters in logging.hpp. This had the unintended effect of making
all compile units pull in the majority of boost::url, and nlohmann::json
as includes.
This commit breaks out boost and json formatters into their own separate
includes.
Tested: Code compiles. Logging changes only.
Change-Id: I6a788533169f10e19130a1910cd3be0cc729b020
Signed-off-by: Ed Tanous <ed@tanous.net>
|
|
Webpack (which is what vue uses to compress its HTML) is capable of
generating hashes of files when it produces the dist files[1].
This gets generated in the form of
<filename>.<hash>.<extension>
This commit attempts to detect these patterns, and enable etag caching
to speed up webui load times. It detects these patterns, grabs the hash
for the file, and returns it in the Etag header[2].
The behavior is implemented such that:
If the file has an etag, the etag header is returned.
If the request has an If-None-Match header, and that header matches,
only 304 is returned.
Tested:
Tests were run on qemu S7106 bmcweb with default error logging level,
and HTTP/2 enabled, along with svg optimization patches.
Run scripts/generate_auth_certificate.py to set up TLS certificates.
(valid TLS certs are required for HTTP caching to work properly in some
browsers).
Load the webui. Note that DOM load takes 1.10 seconds, Load takes 1.10
seconds, and all requests return 200 OK.
Refresh the GUI.
Note that most resources now return 304, and DOM time is reduced to
279 milliseconds and load is reduced to 280 milliseconds. DOM load
(which is what the BMC has control over) is decreased by a factor of
3-4X.
Setting chrome to "Fast 5g" throttling in the network tab shows a more
pronounced difference, 1.28S load time vs 3.96S.
BMC also shows 477KB transferred on the wire, versus 2.3KB
transferred on the wire. This has the potential to significantly
reduce the load on the BMC when the webui refreshes.
[1] https://webpack.js.org/guides/caching/
[2] https://developer.mozilla.org/en-US/docs/Web/HTTP/Headers/ETag
Change-Id: I68aa7ef75533506d98e8fce10bb04a494dc49669
Signed-off-by: Ed Tanous <ed@tanous.net>
|
|
Boost process v2 brings some significant benefits to our launching of
processes[1]. In bmcweb terms:
1. The code is radically simpler, which decreaeses compile times, and
reduces the scope for code scanning tools.
2. The code now uses standard asio pipes instead of inventing its own.
3. Separate compilation.
Tested:
We don't have a lot of unit tests for the virtual media stuff that I can
run, but we do have unit tests for credentials pipe, which in this
change have been ported over, so the feature works. Unit tests are
passing.
[1] https://www.boost.org/doc/libs/1_80_0/doc/html/boost_process/v2.html#boost_process.v2.introduction
Change-Id: Ia20226819d75ff6e492f8852185f0b73e8f5cf83
Signed-off-by: Ed Tanous <ed@tanous.net>
|
|
Currently, the Cookie auth only checks the first cookie header in a
request. This works fine for most things, because a lot of
implementations (browsers) seem to either put the Cookie headers in
alphabetical order, or put them in the order in which they were stored
which in the case of bmcweb, is also alphabetical.
Well, http2 blows this up, because cookies could potentially be in any
order, given the hpack compression techniques, so there's no promise
that Cookie[0] is the Session cookie.
This commit reworks the authentication code to call beasts "equal_range"
getter, which returns the range of all headers that matched. This
allows us to attempt to parse the cookies in whatever order they might
have been received.
The auth routine only tries to log in the first cookie matching
SESSION=, and do not try to handle duplicates, as this might allow
attackers to negate the anti brute force measures by testing multiple
passwords at once
Tested:
With http2 enabled, the UI can now log in more consistently, and in
addition, the HTML redfish pages function more consistently when using
cookie auth.
Redfish service validator passes.
Change-Id: I3a61a5a654f62096ff19cfbfaf0a10f30a1a3605
Signed-off-by: Ed Tanous <ed@tanous.net>
|
|
This feature was created for a time before webpack had a built in proxy,
and to debug the UI required setting specific flags. The webpack proxy
solves this problem in a much better way, by proxying everything.
This commit is one piece in the solving a use after free bug. Removing
this allows us to no longer have to cache the origin header [1], which
is only used in this mode.
Tested: Code compiles.
[1] https://gerrit.openbmc.org/c/openbmc/bmcweb/+/70850
Change-Id: I01d67006e217c0c9fd2db7526c0ec34b0da068f3
Signed-off-by: Ed Tanous <ed@tanous.net>
|
|
This TODO has been in bmcweb for a very long time. Implement it.
W3 sets rules for what security policies apply to which content
types[1]. Reading through this, essentially CSP should only apply to
HTML files.
Tested: Unit tests pass. Webui loads properly. Chrome network window
Shows headers show up as expected.
[1] https://www.w3.org/TR/CSP2/#which-policy-applies
Change-Id: I5467d0373832668763c72a66da2a8872e07bfb58
Signed-off-by: Ed Tanous <ed@tanous.net>
|
|
Most of these were found by breaking every redfish class handler into
its own compile unit:
When that's done, these missing headers become compile errors. We
should just fix them.
In addition, this allows us to enable automatic header checking in
clang-tidy using misc-header-cleaner. Because the compiler can now
"see" all the defines, it no longer tries to remove headers that it
thinks are unused.
[1] https://github.com/openbmc/bmcweb/commit/4fdee9e39e9f03122ee16a6fb251a380681f56ac
Tested: Code compiles.
Change-Id: Ifa27ac4a512362b7ded7cc3068648dc4aea6ad7b
Signed-off-by: Ed Tanous <ed@tanous.net>
|
|
This feature was introduced to manage the operation sync at BMC while
multiple clients manage the BMC.
This feature scope has gone away and it is not a simple code to maintain
as per the growing standards of bmcweb.
This commit removes the feature from this repo.
Tested by: Locks routes are not available anymore
Change-Id: I257225cfb1f43d7d5dadb21a28a2ee5345c5112a
Signed-off-by: Sunitha Harish <sunithaharish04@gmail.com>
Signed-off-by: Ed Tanous <ed@tanous.net>
|
|
This macro came originally from CROW_ENABLE_SSL, and was used as a macro
to optionally compile without openssl being required.
OpenSSL has been pulled into many other dependencies, and has been
functionally required to be included for a long time, so there's no
reason to hold onto this macro.
Remove most uses of the macro, and for the couple functional places the
macro is used, transition to a constexpr if to enable the TLS paths.
This allows a large simplification of code in some places.
Tested: Redfish service validator passes.
Change-Id: Iebd46a68e5e417b6031479e24be3c21bef782f4c
Signed-off-by: Ed Tanous <ed@tanous.net>
|
|
If we include OpenSSL in extern "C" blocks consistently, c++ warnings no
longer appear. This means we can remove the special case from meson.
Tested: Code compiles when built locally on an ubuntu 22.04 system.
Change-Id: I5add4113b32cd88b7fdd874174c845425a7c287a
Signed-off-by: Ed Tanous <ed@tanous.net>
|
|
Several places that call *req.ioService were missing nullptr checks.
Add them, and fix the one case where it might not be filled in.
Tested: With HTTP2 enabled, the following command succeeds.
```
curl -k https://192.168.7.2/redfish/v1/UpdateService/update -F 'UpdateParameters={"Targets":["/redfish/v1/Managers/bmc"]} ;type=application/json' --user "root:0penBmc" -F UpdateFile=@/home/ed/bmcweb/16mb.txt -v -H "Expect:"
```
Change-Id: I81e7944c22f5922d461bf5d231086c7468a16e62
Signed-off-by: Ed Tanous <ed@tanous.net>
|
|
This code used std::remove, which is a mechanism for removing characters
from strings. Clearly it meant std::filesystem::remove(), which removes
files from the filesystem.
Correct it.
Change-Id: I030966203c1682a11c723c596accdf34637dd1ba
Signed-off-by: Ed Tanous <ed@tanous.net>
|
|
boost::asio::buffer_copy returns an integer of the number of values
copied. Some static analysis tools mark that value as nodiscard,
although it should never fail. Audit all uses of buffer_copy, and make
sure that they're using the return value. In theory this should have
no change on the behavior.
Change-Id: I6af39b5347954c2932cf3d4e48e96ff9ae01583a
Signed-off-by: Ed Tanous <ed@tanous.net>
|
|
Despite these objects being called "view" they are still relatively
large, as clang-tidy correctly flags, and we ignore.
Change all function uses to capture by:
const boost::urls::url_view_base&
Which is the base class of all boost URL types, and any class (url,
url_view, etc) is convertible to that base.
Change-Id: I8ee2ea3f4cfba38331303a7e4eb520a2b6f8ba92
Signed-off-by: Ed Tanous <ed@tanous.net>
|
|
Now that we only support string types in the router we no longer need to
build a "Tag" to be used for constructing argument types. Now, we can
just track the number of arguments, which simplifies the code
significantly, and removes the need to convert to and from the tag to
parameter counts.
This in turn deletes a lot of code in the router, removing the need for
tracking tag types.
Tested: Redfish service validator passes. Unit tests pass.
Change-Id: Ide1d665dc1984552681e8c05952b38073d5e32dd
Signed-off-by: Ed Tanous <ed@tanous.net>
|
|
Meson supports unity builds[1] natively. There's no reason to continue
with the pseudo unity build we've been using by putting implementations
in header files.
This commit is the first in a long series of starting to break this up
into smaller compile units, in the hopes of dropping incremental compile
times for developers, and reduce the total per-core memory usage that
gcc requires.
This commit breaks out the run() function from main() and the
constructor of RedfishService from redfish.hpp into their own compile
units. According to tracing, even after broken out, these are still by
far the two longest to compile units in the build.
Tested: Code compiles. Debug build on a 24 core build server results in
a decrease in compile time for compiling just bmcweb from 1m38s to
1m22s.
[1] https://mesonbuild.com/Unity-builds.html
Change-Id: Ibf352e8aba61d64c9a41a7a76e94ab3b5a0dde4b
Signed-off-by: Ed Tanous <ed@tanous.net>
|
|
Clang has new checks for std::move/std::forward correctness, which
catches quite a few "wrong" things where we were making copies of
callback handlers.
Unfortunately, the lambda syntax of
callback{std::forward<Callback>(callback)}
in a capture confuses it, so change usages to
callback = std::forward<Callback>(callback)
to be consistent.
Tested: Redfish service validator passes.
Change-Id: I7a111ec00cf78ecb7d5f5b102c786c1c14d74384
Signed-off-by: Ed Tanous <ed@tanous.net>
|
|
nlohmann::json::dump() is not an easy function to get the call
parameters correct on. We should limit the places we use it.
Luckily, both logging and redfish::messages support printing
json values directly. Use them where appropriate.
Tested: Error logging and out of range calls only of heavily used
messages and logging calls. Inspection only.
Change-Id: I57521d8791dd95250c93e8e3b2d4a959740ac713
Signed-off-by: Ed Tanous <ed@tanous.net>
|
|
inline is not required on member methods. Clang-tidy has a check for
this. Enable the check and fix the two bad usages.
Tested: Code compiles.
Change-Id: I3115b7c0c4005e1082e0005b818fbe6569511f49
Signed-off-by: Ed Tanous <ed@tanous.net>
|
|
clang-18 improves this check so that we can actually use it. Enable it
and fix all violations.
Change-Id: Ibe4ce19c423d447a4cbe593d1abba948362426af
Signed-off-by: Ed Tanous <ed@tanous.net>
|
