postgrex

Summary

Improper Neutralization of Special Elements used in an SQL Command ('SQL Injection') vulnerability in elixir-ecto postgrex allows SQL Injection via the :comment option of Postgrex.stream/4. An attacker who can influence that value can close the comment delimiter with */ and extend the streamed statement with their own clauses, which execute under the connection's role. Ecto exposes the same option through Ecto.Repo.stream/2.

Postgrex appends the comment by concatenating it into the statement text sent in the Parse message, without escaping or rejecting */. The option is validated by comment_not_present!/1 at every other execution point; stream/4 never calls it. Because Parse accepts a single command, the injection is confined to the streamed statement and further statements cannot be chained.

This issue affects postgrex: from 0.19.3 before 0.22.4.

Workarounds

Reject any :comment value containing */ or a null byte before passing it to Postgrex.stream/4 or Ecto.Repo.stream/2. Alternatively, build comments only from trusted values.

Configurations

The application must pass a :comment derived from untrusted input to Postgrex.stream/4 or Ecto.Repo.stream/2. Applications that omit the option or pass only static values are unaffected.

Summary

Improper Neutralization of Special Elements used in an SQL Command ('SQL Injection') vulnerability in elixir-ecto postgrex allows SQL Injection via the :comment option of Postgrex.stream/4. An attacker who can influence that value can close the comment delimiter with */ and extend the streamed statement with their own clauses, which execute under the connection's role. Ecto exposes the same option through Ecto.Repo.stream/2.

Postgrex appends the comment by concatenating it into the statement text sent in the Parse message, without escaping or rejecting */. The option is validated by comment_not_present!/1 at every other execution point; stream/4 never calls it. Because Parse accepts a single command, the injection is confined to the streamed statement and further statements cannot be chained.

This issue affects postgrex: from 0.19.3 before 0.22.4.

Workarounds

Reject any :comment value containing */ or a null byte before passing it to Postgrex.stream/4 or Ecto.Repo.stream/2. Alternatively, build comments only from trusted values.

Configurations

The application must pass a :comment derived from untrusted input to Postgrex.stream/4 or Ecto.Repo.stream/2. Applications that omit the option or pass only static values are unaffected.

Summary

SQL Injection vulnerability in elixir-ecto postgrex allows an attacker who can influence a LISTEN channel name to inject SQL into the reconnect replay query, causing a denial of service of the notification connection.

Postgrex.Notifications sanitizes channel names with quote_channel/1, which doubles double quotes so the name is safe inside a double-quoted identifier. This protects the single-statement LISTEN and UNLISTEN paths. On every (re)connect, however, handle_connect/1 replays all registered channels at once by concatenating their LISTEN statements and wrapping them in a dollar-quoted anonymous code block (DO $$BEGIN ... END$$). quote_channel/1 does not escape the $$ dollar-quote delimiter that opens and closes this block.

The listen/3 guards only reject null bytes and names longer than 63 bytes, so a channel name containing $$ passes validation unchanged. Once such a name is embedded, its $$ prematurely terminates the outer dollar-quoted string and PostgreSQL parses the remainder as additional top-level statements. Because handle_connect/1 runs on every (re)connect, the malformed replay query is rejected each time and the notification connection never re-establishes its subscriptions, silently dropping notifications for every channel sharing that connection.

An application is affected when it passes untrusted input (for example a tenant or user identifier) as a channel name to Postgrex.Notifications.listen/3. The double-quote doubling prevents forming a fully valid injected statement, so arbitrary SQL execution is not possible, but the corrupted query reliably breaks the shared notification connection for all tenants, resulting in denial of service.

This issue affects postgrex: from 0.16.0 before 0.22.3.

Workarounds

Validate channel names before passing untrusted input to Postgrex.Notifications.listen/3. Reject any name that contains the dollar-quote delimiter ($$), or restrict channel names to a safe character set such as alphanumeric characters and underscores.

Summary

Improper Neutralization of Special Elements used in an SQL Command ('SQL Injection') vulnerability in elixir-ecto postgrex allows SQL Injection via the :comment option of Postgrex.stream/4. An attacker who can influence that value can close the comment delimiter with */ and extend the streamed statement with their own clauses, which execute under the connection's role. Ecto exposes the same option through Ecto.Repo.stream/2.

Postgrex appends the comment by concatenating it into the statement text sent in the Parse message, without escaping or rejecting */. The option is validated by comment_not_present!/1 at every other execution point; stream/4 never calls it. Because Parse accepts a single command, the injection is confined to the streamed statement and further statements cannot be chained.

This issue affects postgrex: from 0.19.3 before 0.22.4.

Workarounds

Reject any :comment value containing */ or a null byte before passing it to Postgrex.stream/4 or Ecto.Repo.stream/2. Alternatively, build comments only from trusted values.

Configurations

The application must pass a :comment derived from untrusted input to Postgrex.stream/4 or Ecto.Repo.stream/2. Applications that omit the option or pass only static values are unaffected.

Summary

SQL Injection vulnerability in elixir-ecto postgrex allows an attacker who can influence a LISTEN channel name to inject SQL into the reconnect replay query, causing a denial of service of the notification connection.

Postgrex.Notifications sanitizes channel names with quote_channel/1, which doubles double quotes so the name is safe inside a double-quoted identifier. This protects the single-statement LISTEN and UNLISTEN paths. On every (re)connect, however, handle_connect/1 replays all registered channels at once by concatenating their LISTEN statements and wrapping them in a dollar-quoted anonymous code block (DO $$BEGIN ... END$$). quote_channel/1 does not escape the $$ dollar-quote delimiter that opens and closes this block.

The listen/3 guards only reject null bytes and names longer than 63 bytes, so a channel name containing $$ passes validation unchanged. Once such a name is embedded, its $$ prematurely terminates the outer dollar-quoted string and PostgreSQL parses the remainder as additional top-level statements. Because handle_connect/1 runs on every (re)connect, the malformed replay query is rejected each time and the notification connection never re-establishes its subscriptions, silently dropping notifications for every channel sharing that connection.

An application is affected when it passes untrusted input (for example a tenant or user identifier) as a channel name to Postgrex.Notifications.listen/3. The double-quote doubling prevents forming a fully valid injected statement, so arbitrary SQL execution is not possible, but the corrupted query reliably breaks the shared notification connection for all tenants, resulting in denial of service.

This issue affects postgrex: from 0.16.0 before 0.22.3.

Workarounds

Validate channel names before passing untrusted input to Postgrex.Notifications.listen/3. Reject any name that contains the dollar-quote delimiter ($$), or restrict channel names to a safe character set such as alphanumeric characters and underscores.

Summary

SQL injection in Postgrex.Notifications.listen/3: the channel argument is interpolated straight into LISTEN "..." / UNLISTEN "..." without escaping the " character. Any caller that lets a user influence the channel name (e.g. a pub/sub bridge that uses a tenant id or topic slug as the channel) and the name is not sanitized can execute arbitrary SQL on the notifications connection.

Only those using the Postgrex.Notifications directly or via a dependency, with a non-sanitized channel name are vulnerable. Ecto does not use Postgrex.Notifications by default, but you must validate if any other dependency does.

Details

PostgreSQL escapes a " inside a quoted identifier by doubling it to "". Postgrex doesn't do that doubling, so a " inside channel closes the identifier early and everything after it is parsed as SQL on the same connection.

The notifications connection runs as whatever DB role the app is configured with. Unlike the Postgrex.query/4 API, there's no extended-protocol guard here: LISTEN/UNLISTEN are sent as simple queries, so multi-statement payloads work and the attacker can chain ; CREATE TABLE …, ; DROP …, ; CREATE ROLE …, etc.

Impact

SQL injection on the notifications connection. Affects any application that calls Postgrex.Notifications.listen/3 (or unlisten/3) with a channel name derived from untrusted input. Executes as the configured DB user with no protocol-level limit on what can be chained, so the realistic blast radius is read/modify/destroy on any data the app's DB role can reach.

Fix

Upgrade to latest Postgrex v0.22.2 or later. Altenratively, sanitize any user input given as channel name by making sure it doesn't include quotes as well as null bytes (not strictly required, but recommended):

if String.contains?(channel_name, ["\"", "\0"]) do
  raise "bad channel name"
end

Summary

Improper Neutralization of Special Elements used in an SQL Command ('SQL Injection') vulnerability in elixir-ecto postgrex allows SQL Injection via the :comment option of Postgrex.stream/4. An attacker who can influence that value can close the comment delimiter with */ and extend the streamed statement with their own clauses, which execute under the connection's role. Ecto exposes the same option through Ecto.Repo.stream/2.

Postgrex appends the comment by concatenating it into the statement text sent in the Parse message, without escaping or rejecting */. The option is validated by comment_not_present!/1 at every other execution point; stream/4 never calls it. Because Parse accepts a single command, the injection is confined to the streamed statement and further statements cannot be chained.

This issue affects postgrex: from 0.19.3 before 0.22.4.

Workarounds

Reject any :comment value containing */ or a null byte before passing it to Postgrex.stream/4 or Ecto.Repo.stream/2. Alternatively, build comments only from trusted values.

Configurations

The application must pass a :comment derived from untrusted input to Postgrex.stream/4 or Ecto.Repo.stream/2. Applications that omit the option or pass only static values are unaffected.

Summary

SQL Injection vulnerability in elixir-ecto postgrex allows an attacker who can influence a LISTEN channel name to inject SQL into the reconnect replay query, causing a denial of service of the notification connection.

Postgrex.Notifications sanitizes channel names with quote_channel/1, which doubles double quotes so the name is safe inside a double-quoted identifier. This protects the single-statement LISTEN and UNLISTEN paths. On every (re)connect, however, handle_connect/1 replays all registered channels at once by concatenating their LISTEN statements and wrapping them in a dollar-quoted anonymous code block (DO $$BEGIN ... END$$). quote_channel/1 does not escape the $$ dollar-quote delimiter that opens and closes this block.

The listen/3 guards only reject null bytes and names longer than 63 bytes, so a channel name containing $$ passes validation unchanged. Once such a name is embedded, its $$ prematurely terminates the outer dollar-quoted string and PostgreSQL parses the remainder as additional top-level statements. Because handle_connect/1 runs on every (re)connect, the malformed replay query is rejected each time and the notification connection never re-establishes its subscriptions, silently dropping notifications for every channel sharing that connection.

An application is affected when it passes untrusted input (for example a tenant or user identifier) as a channel name to Postgrex.Notifications.listen/3. The double-quote doubling prevents forming a fully valid injected statement, so arbitrary SQL execution is not possible, but the corrupted query reliably breaks the shared notification connection for all tenants, resulting in denial of service.

This issue affects postgrex: from 0.16.0 before 0.22.3.

Workarounds

Validate channel names before passing untrusted input to Postgrex.Notifications.listen/3. Reject any name that contains the dollar-quote delimiter ($$), or restrict channel names to a safe character set such as alphanumeric characters and underscores.

Summary

SQL injection in Postgrex.Notifications.listen/3: the channel argument is interpolated straight into LISTEN "..." / UNLISTEN "..." without escaping the " character. Any caller that lets a user influence the channel name (e.g. a pub/sub bridge that uses a tenant id or topic slug as the channel) and the name is not sanitized can execute arbitrary SQL on the notifications connection.

Only those using the Postgrex.Notifications directly or via a dependency, with a non-sanitized channel name are vulnerable. Ecto does not use Postgrex.Notifications by default, but you must validate if any other dependency does.

Details

PostgreSQL escapes a " inside a quoted identifier by doubling it to "". Postgrex doesn't do that doubling, so a " inside channel closes the identifier early and everything after it is parsed as SQL on the same connection.

The notifications connection runs as whatever DB role the app is configured with. Unlike the Postgrex.query/4 API, there's no extended-protocol guard here: LISTEN/UNLISTEN are sent as simple queries, so multi-statement payloads work and the attacker can chain ; CREATE TABLE …, ; DROP …, ; CREATE ROLE …, etc.

Impact

SQL injection on the notifications connection. Affects any application that calls Postgrex.Notifications.listen/3 (or unlisten/3) with a channel name derived from untrusted input. Executes as the configured DB user with no protocol-level limit on what can be chained, so the realistic blast radius is read/modify/destroy on any data the app's DB role can reach.

Fix

Upgrade to latest Postgrex v0.22.2 or later. Altenratively, sanitize any user input given as channel name by making sure it doesn't include quotes as well as null bytes (not strictly required, but recommended):

if String.contains?(channel_name, ["\"", "\0"]) do
  raise "bad channel name"
end

Summary

Improper Neutralization of Special Elements used in an SQL Command ('SQL Injection') vulnerability in elixir-ecto postgrex allows SQL Injection via the :comment option of Postgrex.stream/4. An attacker who can influence that value can close the comment delimiter with */ and extend the streamed statement with their own clauses, which execute under the connection's role. Ecto exposes the same option through Ecto.Repo.stream/2.

Postgrex appends the comment by concatenating it into the statement text sent in the Parse message, without escaping or rejecting */. The option is validated by comment_not_present!/1 at every other execution point; stream/4 never calls it. Because Parse accepts a single command, the injection is confined to the streamed statement and further statements cannot be chained.

This issue affects postgrex: from 0.19.3 before 0.22.4.

Workarounds

Reject any :comment value containing */ or a null byte before passing it to Postgrex.stream/4 or Ecto.Repo.stream/2. Alternatively, build comments only from trusted values.

Configurations

The application must pass a :comment derived from untrusted input to Postgrex.stream/4 or Ecto.Repo.stream/2. Applications that omit the option or pass only static values are unaffected.

Summary

SQL Injection vulnerability in elixir-ecto postgrex allows an attacker who can influence a LISTEN channel name to inject SQL into the reconnect replay query, causing a denial of service of the notification connection.

Postgrex.Notifications sanitizes channel names with quote_channel/1, which doubles double quotes so the name is safe inside a double-quoted identifier. This protects the single-statement LISTEN and UNLISTEN paths. On every (re)connect, however, handle_connect/1 replays all registered channels at once by concatenating their LISTEN statements and wrapping them in a dollar-quoted anonymous code block (DO $$BEGIN ... END$$). quote_channel/1 does not escape the $$ dollar-quote delimiter that opens and closes this block.

The listen/3 guards only reject null bytes and names longer than 63 bytes, so a channel name containing $$ passes validation unchanged. Once such a name is embedded, its $$ prematurely terminates the outer dollar-quoted string and PostgreSQL parses the remainder as additional top-level statements. Because handle_connect/1 runs on every (re)connect, the malformed replay query is rejected each time and the notification connection never re-establishes its subscriptions, silently dropping notifications for every channel sharing that connection.

An application is affected when it passes untrusted input (for example a tenant or user identifier) as a channel name to Postgrex.Notifications.listen/3. The double-quote doubling prevents forming a fully valid injected statement, so arbitrary SQL execution is not possible, but the corrupted query reliably breaks the shared notification connection for all tenants, resulting in denial of service.

This issue affects postgrex: from 0.16.0 before 0.22.3.

Workarounds

Validate channel names before passing untrusted input to Postgrex.Notifications.listen/3. Reject any name that contains the dollar-quote delimiter ($$), or restrict channel names to a safe character set such as alphanumeric characters and underscores.

Summary

SQL injection in Postgrex.Notifications.listen/3: the channel argument is interpolated straight into LISTEN "..." / UNLISTEN "..." without escaping the " character. Any caller that lets a user influence the channel name (e.g. a pub/sub bridge that uses a tenant id or topic slug as the channel) and the name is not sanitized can execute arbitrary SQL on the notifications connection.

Only those using the Postgrex.Notifications directly or via a dependency, with a non-sanitized channel name are vulnerable. Ecto does not use Postgrex.Notifications by default, but you must validate if any other dependency does.

Details

PostgreSQL escapes a " inside a quoted identifier by doubling it to "". Postgrex doesn't do that doubling, so a " inside channel closes the identifier early and everything after it is parsed as SQL on the same connection.

The notifications connection runs as whatever DB role the app is configured with. Unlike the Postgrex.query/4 API, there's no extended-protocol guard here: LISTEN/UNLISTEN are sent as simple queries, so multi-statement payloads work and the attacker can chain ; CREATE TABLE …, ; DROP …, ; CREATE ROLE …, etc.

Impact

SQL injection on the notifications connection. Affects any application that calls Postgrex.Notifications.listen/3 (or unlisten/3) with a channel name derived from untrusted input. Executes as the configured DB user with no protocol-level limit on what can be chained, so the realistic blast radius is read/modify/destroy on any data the app's DB role can reach.

Fix

Upgrade to latest Postgrex v0.22.2 or later. Altenratively, sanitize any user input given as channel name by making sure it doesn't include quotes as well as null bytes (not strictly required, but recommended):

if String.contains?(channel_name, ["\"", "\0"]) do
  raise "bad channel name"
end

Summary

Improper Neutralization of Special Elements used in an SQL Command ('SQL Injection') vulnerability in elixir-ecto postgrex allows SQL Injection via the :comment option of Postgrex.stream/4. An attacker who can influence that value can close the comment delimiter with */ and extend the streamed statement with their own clauses, which execute under the connection's role. Ecto exposes the same option through Ecto.Repo.stream/2.

Postgrex appends the comment by concatenating it into the statement text sent in the Parse message, without escaping or rejecting */. The option is validated by comment_not_present!/1 at every other execution point; stream/4 never calls it. Because Parse accepts a single command, the injection is confined to the streamed statement and further statements cannot be chained.

This issue affects postgrex: from 0.19.3 before 0.22.4.

Workarounds

Reject any :comment value containing */ or a null byte before passing it to Postgrex.stream/4 or Ecto.Repo.stream/2. Alternatively, build comments only from trusted values.

Configurations

The application must pass a :comment derived from untrusted input to Postgrex.stream/4 or Ecto.Repo.stream/2. Applications that omit the option or pass only static values are unaffected.

Summary

SQL Injection vulnerability in elixir-ecto postgrex allows an attacker who can influence a LISTEN channel name to inject SQL into the reconnect replay query, causing a denial of service of the notification connection.

Postgrex.Notifications sanitizes channel names with quote_channel/1, which doubles double quotes so the name is safe inside a double-quoted identifier. This protects the single-statement LISTEN and UNLISTEN paths. On every (re)connect, however, handle_connect/1 replays all registered channels at once by concatenating their LISTEN statements and wrapping them in a dollar-quoted anonymous code block (DO $$BEGIN ... END$$). quote_channel/1 does not escape the $$ dollar-quote delimiter that opens and closes this block.

The listen/3 guards only reject null bytes and names longer than 63 bytes, so a channel name containing $$ passes validation unchanged. Once such a name is embedded, its $$ prematurely terminates the outer dollar-quoted string and PostgreSQL parses the remainder as additional top-level statements. Because handle_connect/1 runs on every (re)connect, the malformed replay query is rejected each time and the notification connection never re-establishes its subscriptions, silently dropping notifications for every channel sharing that connection.

An application is affected when it passes untrusted input (for example a tenant or user identifier) as a channel name to Postgrex.Notifications.listen/3. The double-quote doubling prevents forming a fully valid injected statement, so arbitrary SQL execution is not possible, but the corrupted query reliably breaks the shared notification connection for all tenants, resulting in denial of service.

This issue affects postgrex: from 0.16.0 before 0.22.3.

Workarounds

Validate channel names before passing untrusted input to Postgrex.Notifications.listen/3. Reject any name that contains the dollar-quote delimiter ($$), or restrict channel names to a safe character set such as alphanumeric characters and underscores.

Summary

SQL injection in Postgrex.Notifications.listen/3: the channel argument is interpolated straight into LISTEN "..." / UNLISTEN "..." without escaping the " character. Any caller that lets a user influence the channel name (e.g. a pub/sub bridge that uses a tenant id or topic slug as the channel) and the name is not sanitized can execute arbitrary SQL on the notifications connection.

Only those using the Postgrex.Notifications directly or via a dependency, with a non-sanitized channel name are vulnerable. Ecto does not use Postgrex.Notifications by default, but you must validate if any other dependency does.

Details

PostgreSQL escapes a " inside a quoted identifier by doubling it to "". Postgrex doesn't do that doubling, so a " inside channel closes the identifier early and everything after it is parsed as SQL on the same connection.

The notifications connection runs as whatever DB role the app is configured with. Unlike the Postgrex.query/4 API, there's no extended-protocol guard here: LISTEN/UNLISTEN are sent as simple queries, so multi-statement payloads work and the attacker can chain ; CREATE TABLE …, ; DROP …, ; CREATE ROLE …, etc.

Impact

SQL injection on the notifications connection. Affects any application that calls Postgrex.Notifications.listen/3 (or unlisten/3) with a channel name derived from untrusted input. Executes as the configured DB user with no protocol-level limit on what can be chained, so the realistic blast radius is read/modify/destroy on any data the app's DB role can reach.

Fix

Upgrade to latest Postgrex v0.22.2 or later. Altenratively, sanitize any user input given as channel name by making sure it doesn't include quotes as well as null bytes (not strictly required, but recommended):

if String.contains?(channel_name, ["\"", "\0"]) do
  raise "bad channel name"
end

Summary

Improper Neutralization of Special Elements used in an SQL Command ('SQL Injection') vulnerability in elixir-ecto postgrex allows SQL Injection via the :comment option of Postgrex.stream/4. An attacker who can influence that value can close the comment delimiter with */ and extend the streamed statement with their own clauses, which execute under the connection's role. Ecto exposes the same option through Ecto.Repo.stream/2.

Postgrex appends the comment by concatenating it into the statement text sent in the Parse message, without escaping or rejecting */. The option is validated by comment_not_present!/1 at every other execution point; stream/4 never calls it. Because Parse accepts a single command, the injection is confined to the streamed statement and further statements cannot be chained.

This issue affects postgrex: from 0.19.3 before 0.22.4.

Workarounds

Reject any :comment value containing */ or a null byte before passing it to Postgrex.stream/4 or Ecto.Repo.stream/2. Alternatively, build comments only from trusted values.

Configurations

The application must pass a :comment derived from untrusted input to Postgrex.stream/4 or Ecto.Repo.stream/2. Applications that omit the option or pass only static values are unaffected.

Summary

SQL Injection vulnerability in elixir-ecto postgrex allows an attacker who can influence a LISTEN channel name to inject SQL into the reconnect replay query, causing a denial of service of the notification connection.

Postgrex.Notifications sanitizes channel names with quote_channel/1, which doubles double quotes so the name is safe inside a double-quoted identifier. This protects the single-statement LISTEN and UNLISTEN paths. On every (re)connect, however, handle_connect/1 replays all registered channels at once by concatenating their LISTEN statements and wrapping them in a dollar-quoted anonymous code block (DO $$BEGIN ... END$$). quote_channel/1 does not escape the $$ dollar-quote delimiter that opens and closes this block.

The listen/3 guards only reject null bytes and names longer than 63 bytes, so a channel name containing $$ passes validation unchanged. Once such a name is embedded, its $$ prematurely terminates the outer dollar-quoted string and PostgreSQL parses the remainder as additional top-level statements. Because handle_connect/1 runs on every (re)connect, the malformed replay query is rejected each time and the notification connection never re-establishes its subscriptions, silently dropping notifications for every channel sharing that connection.

An application is affected when it passes untrusted input (for example a tenant or user identifier) as a channel name to Postgrex.Notifications.listen/3. The double-quote doubling prevents forming a fully valid injected statement, so arbitrary SQL execution is not possible, but the corrupted query reliably breaks the shared notification connection for all tenants, resulting in denial of service.

This issue affects postgrex: from 0.16.0 before 0.22.3.

Workarounds

Validate channel names before passing untrusted input to Postgrex.Notifications.listen/3. Reject any name that contains the dollar-quote delimiter ($$), or restrict channel names to a safe character set such as alphanumeric characters and underscores.

Summary

SQL injection in Postgrex.Notifications.listen/3: the channel argument is interpolated straight into LISTEN "..." / UNLISTEN "..." without escaping the " character. Any caller that lets a user influence the channel name (e.g. a pub/sub bridge that uses a tenant id or topic slug as the channel) and the name is not sanitized can execute arbitrary SQL on the notifications connection.

Only those using the Postgrex.Notifications directly or via a dependency, with a non-sanitized channel name are vulnerable. Ecto does not use Postgrex.Notifications by default, but you must validate if any other dependency does.

Details

PostgreSQL escapes a " inside a quoted identifier by doubling it to "". Postgrex doesn't do that doubling, so a " inside channel closes the identifier early and everything after it is parsed as SQL on the same connection.

The notifications connection runs as whatever DB role the app is configured with. Unlike the Postgrex.query/4 API, there's no extended-protocol guard here: LISTEN/UNLISTEN are sent as simple queries, so multi-statement payloads work and the attacker can chain ; CREATE TABLE …, ; DROP …, ; CREATE ROLE …, etc.

Impact

SQL injection on the notifications connection. Affects any application that calls Postgrex.Notifications.listen/3 (or unlisten/3) with a channel name derived from untrusted input. Executes as the configured DB user with no protocol-level limit on what can be chained, so the realistic blast radius is read/modify/destroy on any data the app's DB role can reach.

Fix

Upgrade to latest Postgrex v0.22.2 or later. Altenratively, sanitize any user input given as channel name by making sure it doesn't include quotes as well as null bytes (not strictly required, but recommended):

if String.contains?(channel_name, ["\"", "\0"]) do
  raise "bad channel name"
end

Summary

Improper Neutralization of Special Elements used in an SQL Command ('SQL Injection') vulnerability in elixir-ecto postgrex allows SQL Injection via the :comment option of Postgrex.stream/4. An attacker who can influence that value can close the comment delimiter with */ and extend the streamed statement with their own clauses, which execute under the connection's role. Ecto exposes the same option through Ecto.Repo.stream/2.

Postgrex appends the comment by concatenating it into the statement text sent in the Parse message, without escaping or rejecting */. The option is validated by comment_not_present!/1 at every other execution point; stream/4 never calls it. Because Parse accepts a single command, the injection is confined to the streamed statement and further statements cannot be chained.

This issue affects postgrex: from 0.19.3 before 0.22.4.

Workarounds

Reject any :comment value containing */ or a null byte before passing it to Postgrex.stream/4 or Ecto.Repo.stream/2. Alternatively, build comments only from trusted values.

Configurations

The application must pass a :comment derived from untrusted input to Postgrex.stream/4 or Ecto.Repo.stream/2. Applications that omit the option or pass only static values are unaffected.

Summary

SQL Injection vulnerability in elixir-ecto postgrex allows an attacker who can influence a LISTEN channel name to inject SQL into the reconnect replay query, causing a denial of service of the notification connection.

Postgrex.Notifications sanitizes channel names with quote_channel/1, which doubles double quotes so the name is safe inside a double-quoted identifier. This protects the single-statement LISTEN and UNLISTEN paths. On every (re)connect, however, handle_connect/1 replays all registered channels at once by concatenating their LISTEN statements and wrapping them in a dollar-quoted anonymous code block (DO $$BEGIN ... END$$). quote_channel/1 does not escape the $$ dollar-quote delimiter that opens and closes this block.

The listen/3 guards only reject null bytes and names longer than 63 bytes, so a channel name containing $$ passes validation unchanged. Once such a name is embedded, its $$ prematurely terminates the outer dollar-quoted string and PostgreSQL parses the remainder as additional top-level statements. Because handle_connect/1 runs on every (re)connect, the malformed replay query is rejected each time and the notification connection never re-establishes its subscriptions, silently dropping notifications for every channel sharing that connection.

An application is affected when it passes untrusted input (for example a tenant or user identifier) as a channel name to Postgrex.Notifications.listen/3. The double-quote doubling prevents forming a fully valid injected statement, so arbitrary SQL execution is not possible, but the corrupted query reliably breaks the shared notification connection for all tenants, resulting in denial of service.

This issue affects postgrex: from 0.16.0 before 0.22.3.

Workarounds

Validate channel names before passing untrusted input to Postgrex.Notifications.listen/3. Reject any name that contains the dollar-quote delimiter ($$), or restrict channel names to a safe character set such as alphanumeric characters and underscores.

Summary

SQL injection in Postgrex.Notifications.listen/3: the channel argument is interpolated straight into LISTEN "..." / UNLISTEN "..." without escaping the " character. Any caller that lets a user influence the channel name (e.g. a pub/sub bridge that uses a tenant id or topic slug as the channel) and the name is not sanitized can execute arbitrary SQL on the notifications connection.

Only those using the Postgrex.Notifications directly or via a dependency, with a non-sanitized channel name are vulnerable. Ecto does not use Postgrex.Notifications by default, but you must validate if any other dependency does.

Details

PostgreSQL escapes a " inside a quoted identifier by doubling it to "". Postgrex doesn't do that doubling, so a " inside channel closes the identifier early and everything after it is parsed as SQL on the same connection.

The notifications connection runs as whatever DB role the app is configured with. Unlike the Postgrex.query/4 API, there's no extended-protocol guard here: LISTEN/UNLISTEN are sent as simple queries, so multi-statement payloads work and the attacker can chain ; CREATE TABLE …, ; DROP …, ; CREATE ROLE …, etc.

Impact

SQL injection on the notifications connection. Affects any application that calls Postgrex.Notifications.listen/3 (or unlisten/3) with a channel name derived from untrusted input. Executes as the configured DB user with no protocol-level limit on what can be chained, so the realistic blast radius is read/modify/destroy on any data the app's DB role can reach.

Fix

Upgrade to latest Postgrex v0.22.2 or later. Altenratively, sanitize any user input given as channel name by making sure it doesn't include quotes as well as null bytes (not strictly required, but recommended):

if String.contains?(channel_name, ["\"", "\0"]) do
  raise "bad channel name"
end

Summary

SQL Injection vulnerability in elixir-ecto postgrex allows an attacker who can influence a LISTEN channel name to inject SQL into the reconnect replay query, causing a denial of service of the notification connection.

Postgrex.Notifications sanitizes channel names with quote_channel/1, which doubles double quotes so the name is safe inside a double-quoted identifier. This protects the single-statement LISTEN and UNLISTEN paths. On every (re)connect, however, handle_connect/1 replays all registered channels at once by concatenating their LISTEN statements and wrapping them in a dollar-quoted anonymous code block (DO $$BEGIN ... END$$). quote_channel/1 does not escape the $$ dollar-quote delimiter that opens and closes this block.

The listen/3 guards only reject null bytes and names longer than 63 bytes, so a channel name containing $$ passes validation unchanged. Once such a name is embedded, its $$ prematurely terminates the outer dollar-quoted string and PostgreSQL parses the remainder as additional top-level statements. Because handle_connect/1 runs on every (re)connect, the malformed replay query is rejected each time and the notification connection never re-establishes its subscriptions, silently dropping notifications for every channel sharing that connection.

An application is affected when it passes untrusted input (for example a tenant or user identifier) as a channel name to Postgrex.Notifications.listen/3. The double-quote doubling prevents forming a fully valid injected statement, so arbitrary SQL execution is not possible, but the corrupted query reliably breaks the shared notification connection for all tenants, resulting in denial of service.

This issue affects postgrex: from 0.16.0 before 0.22.3.

Workarounds

Validate channel names before passing untrusted input to Postgrex.Notifications.listen/3. Reject any name that contains the dollar-quote delimiter ($$), or restrict channel names to a safe character set such as alphanumeric characters and underscores.

Summary

SQL injection in Postgrex.Notifications.listen/3: the channel argument is interpolated straight into LISTEN "..." / UNLISTEN "..." without escaping the " character. Any caller that lets a user influence the channel name (e.g. a pub/sub bridge that uses a tenant id or topic slug as the channel) and the name is not sanitized can execute arbitrary SQL on the notifications connection.

Only those using the Postgrex.Notifications directly or via a dependency, with a non-sanitized channel name are vulnerable. Ecto does not use Postgrex.Notifications by default, but you must validate if any other dependency does.

Details

PostgreSQL escapes a " inside a quoted identifier by doubling it to "". Postgrex doesn't do that doubling, so a " inside channel closes the identifier early and everything after it is parsed as SQL on the same connection.

The notifications connection runs as whatever DB role the app is configured with. Unlike the Postgrex.query/4 API, there's no extended-protocol guard here: LISTEN/UNLISTEN are sent as simple queries, so multi-statement payloads work and the attacker can chain ; CREATE TABLE …, ; DROP …, ; CREATE ROLE …, etc.

Impact

SQL injection on the notifications connection. Affects any application that calls Postgrex.Notifications.listen/3 (or unlisten/3) with a channel name derived from untrusted input. Executes as the configured DB user with no protocol-level limit on what can be chained, so the realistic blast radius is read/modify/destroy on any data the app's DB role can reach.

Fix

Upgrade to latest Postgrex v0.22.2 or later. Altenratively, sanitize any user input given as channel name by making sure it doesn't include quotes as well as null bytes (not strictly required, but recommended):

if String.contains?(channel_name, ["\"", "\0"]) do
  raise "bad channel name"
end

Summary

SQL Injection vulnerability in elixir-ecto postgrex allows an attacker who can influence a LISTEN channel name to inject SQL into the reconnect replay query, causing a denial of service of the notification connection.

Postgrex.Notifications sanitizes channel names with quote_channel/1, which doubles double quotes so the name is safe inside a double-quoted identifier. This protects the single-statement LISTEN and UNLISTEN paths. On every (re)connect, however, handle_connect/1 replays all registered channels at once by concatenating their LISTEN statements and wrapping them in a dollar-quoted anonymous code block (DO $$BEGIN ... END$$). quote_channel/1 does not escape the $$ dollar-quote delimiter that opens and closes this block.

The listen/3 guards only reject null bytes and names longer than 63 bytes, so a channel name containing $$ passes validation unchanged. Once such a name is embedded, its $$ prematurely terminates the outer dollar-quoted string and PostgreSQL parses the remainder as additional top-level statements. Because handle_connect/1 runs on every (re)connect, the malformed replay query is rejected each time and the notification connection never re-establishes its subscriptions, silently dropping notifications for every channel sharing that connection.

An application is affected when it passes untrusted input (for example a tenant or user identifier) as a channel name to Postgrex.Notifications.listen/3. The double-quote doubling prevents forming a fully valid injected statement, so arbitrary SQL execution is not possible, but the corrupted query reliably breaks the shared notification connection for all tenants, resulting in denial of service.

This issue affects postgrex: from 0.16.0 before 0.22.3.

Workarounds

Validate channel names before passing untrusted input to Postgrex.Notifications.listen/3. Reject any name that contains the dollar-quote delimiter ($$), or restrict channel names to a safe character set such as alphanumeric characters and underscores.

Summary

SQL injection in Postgrex.Notifications.listen/3: the channel argument is interpolated straight into LISTEN "..." / UNLISTEN "..." without escaping the " character. Any caller that lets a user influence the channel name (e.g. a pub/sub bridge that uses a tenant id or topic slug as the channel) and the name is not sanitized can execute arbitrary SQL on the notifications connection.

Only those using the Postgrex.Notifications directly or via a dependency, with a non-sanitized channel name are vulnerable. Ecto does not use Postgrex.Notifications by default, but you must validate if any other dependency does.

Details

PostgreSQL escapes a " inside a quoted identifier by doubling it to "". Postgrex doesn't do that doubling, so a " inside channel closes the identifier early and everything after it is parsed as SQL on the same connection.

The notifications connection runs as whatever DB role the app is configured with. Unlike the Postgrex.query/4 API, there's no extended-protocol guard here: LISTEN/UNLISTEN are sent as simple queries, so multi-statement payloads work and the attacker can chain ; CREATE TABLE …, ; DROP …, ; CREATE ROLE …, etc.

Impact

SQL injection on the notifications connection. Affects any application that calls Postgrex.Notifications.listen/3 (or unlisten/3) with a channel name derived from untrusted input. Executes as the configured DB user with no protocol-level limit on what can be chained, so the realistic blast radius is read/modify/destroy on any data the app's DB role can reach.

Fix

Upgrade to latest Postgrex v0.22.2 or later. Altenratively, sanitize any user input given as channel name by making sure it doesn't include quotes as well as null bytes (not strictly required, but recommended):

if String.contains?(channel_name, ["\"", "\0"]) do
  raise "bad channel name"
end

Summary

SQL Injection vulnerability in elixir-ecto postgrex allows an attacker who can influence a LISTEN channel name to inject SQL into the reconnect replay query, causing a denial of service of the notification connection.

Postgrex.Notifications sanitizes channel names with quote_channel/1, which doubles double quotes so the name is safe inside a double-quoted identifier. This protects the single-statement LISTEN and UNLISTEN paths. On every (re)connect, however, handle_connect/1 replays all registered channels at once by concatenating their LISTEN statements and wrapping them in a dollar-quoted anonymous code block (DO $$BEGIN ... END$$). quote_channel/1 does not escape the $$ dollar-quote delimiter that opens and closes this block.

The listen/3 guards only reject null bytes and names longer than 63 bytes, so a channel name containing $$ passes validation unchanged. Once such a name is embedded, its $$ prematurely terminates the outer dollar-quoted string and PostgreSQL parses the remainder as additional top-level statements. Because handle_connect/1 runs on every (re)connect, the malformed replay query is rejected each time and the notification connection never re-establishes its subscriptions, silently dropping notifications for every channel sharing that connection.

An application is affected when it passes untrusted input (for example a tenant or user identifier) as a channel name to Postgrex.Notifications.listen/3. The double-quote doubling prevents forming a fully valid injected statement, so arbitrary SQL execution is not possible, but the corrupted query reliably breaks the shared notification connection for all tenants, resulting in denial of service.

This issue affects postgrex: from 0.16.0 before 0.22.3.

Workarounds

Validate channel names before passing untrusted input to Postgrex.Notifications.listen/3. Reject any name that contains the dollar-quote delimiter ($$), or restrict channel names to a safe character set such as alphanumeric characters and underscores.

Summary

SQL injection in Postgrex.Notifications.listen/3: the channel argument is interpolated straight into LISTEN "..." / UNLISTEN "..." without escaping the " character. Any caller that lets a user influence the channel name (e.g. a pub/sub bridge that uses a tenant id or topic slug as the channel) and the name is not sanitized can execute arbitrary SQL on the notifications connection.

Only those using the Postgrex.Notifications directly or via a dependency, with a non-sanitized channel name are vulnerable. Ecto does not use Postgrex.Notifications by default, but you must validate if any other dependency does.

Details

PostgreSQL escapes a " inside a quoted identifier by doubling it to "". Postgrex doesn't do that doubling, so a " inside channel closes the identifier early and everything after it is parsed as SQL on the same connection.

The notifications connection runs as whatever DB role the app is configured with. Unlike the Postgrex.query/4 API, there's no extended-protocol guard here: LISTEN/UNLISTEN are sent as simple queries, so multi-statement payloads work and the attacker can chain ; CREATE TABLE …, ; DROP …, ; CREATE ROLE …, etc.

Impact

SQL injection on the notifications connection. Affects any application that calls Postgrex.Notifications.listen/3 (or unlisten/3) with a channel name derived from untrusted input. Executes as the configured DB user with no protocol-level limit on what can be chained, so the realistic blast radius is read/modify/destroy on any data the app's DB role can reach.

Fix

Upgrade to latest Postgrex v0.22.2 or later. Altenratively, sanitize any user input given as channel name by making sure it doesn't include quotes as well as null bytes (not strictly required, but recommended):

if String.contains?(channel_name, ["\"", "\0"]) do
  raise "bad channel name"
end

Summary

SQL Injection vulnerability in elixir-ecto postgrex allows an attacker who can influence a LISTEN channel name to inject SQL into the reconnect replay query, causing a denial of service of the notification connection.

Postgrex.Notifications sanitizes channel names with quote_channel/1, which doubles double quotes so the name is safe inside a double-quoted identifier. This protects the single-statement LISTEN and UNLISTEN paths. On every (re)connect, however, handle_connect/1 replays all registered channels at once by concatenating their LISTEN statements and wrapping them in a dollar-quoted anonymous code block (DO $$BEGIN ... END$$). quote_channel/1 does not escape the $$ dollar-quote delimiter that opens and closes this block.

The listen/3 guards only reject null bytes and names longer than 63 bytes, so a channel name containing $$ passes validation unchanged. Once such a name is embedded, its $$ prematurely terminates the outer dollar-quoted string and PostgreSQL parses the remainder as additional top-level statements. Because handle_connect/1 runs on every (re)connect, the malformed replay query is rejected each time and the notification connection never re-establishes its subscriptions, silently dropping notifications for every channel sharing that connection.

An application is affected when it passes untrusted input (for example a tenant or user identifier) as a channel name to Postgrex.Notifications.listen/3. The double-quote doubling prevents forming a fully valid injected statement, so arbitrary SQL execution is not possible, but the corrupted query reliably breaks the shared notification connection for all tenants, resulting in denial of service.

This issue affects postgrex: from 0.16.0 before 0.22.3.

Workarounds

Validate channel names before passing untrusted input to Postgrex.Notifications.listen/3. Reject any name that contains the dollar-quote delimiter ($$), or restrict channel names to a safe character set such as alphanumeric characters and underscores.

Summary

SQL injection in Postgrex.Notifications.listen/3: the channel argument is interpolated straight into LISTEN "..." / UNLISTEN "..." without escaping the " character. Any caller that lets a user influence the channel name (e.g. a pub/sub bridge that uses a tenant id or topic slug as the channel) and the name is not sanitized can execute arbitrary SQL on the notifications connection.

Only those using the Postgrex.Notifications directly or via a dependency, with a non-sanitized channel name are vulnerable. Ecto does not use Postgrex.Notifications by default, but you must validate if any other dependency does.

Details

PostgreSQL escapes a " inside a quoted identifier by doubling it to "". Postgrex doesn't do that doubling, so a " inside channel closes the identifier early and everything after it is parsed as SQL on the same connection.

The notifications connection runs as whatever DB role the app is configured with. Unlike the Postgrex.query/4 API, there's no extended-protocol guard here: LISTEN/UNLISTEN are sent as simple queries, so multi-statement payloads work and the attacker can chain ; CREATE TABLE …, ; DROP …, ; CREATE ROLE …, etc.

Impact

SQL injection on the notifications connection. Affects any application that calls Postgrex.Notifications.listen/3 (or unlisten/3) with a channel name derived from untrusted input. Executes as the configured DB user with no protocol-level limit on what can be chained, so the realistic blast radius is read/modify/destroy on any data the app's DB role can reach.

Fix

Upgrade to latest Postgrex v0.22.2 or later. Altenratively, sanitize any user input given as channel name by making sure it doesn't include quotes as well as null bytes (not strictly required, but recommended):

if String.contains?(channel_name, ["\"", "\0"]) do
  raise "bad channel name"
end

Summary

SQL Injection vulnerability in elixir-ecto postgrex allows an attacker who can influence a LISTEN channel name to inject SQL into the reconnect replay query, causing a denial of service of the notification connection.

Postgrex.Notifications sanitizes channel names with quote_channel/1, which doubles double quotes so the name is safe inside a double-quoted identifier. This protects the single-statement LISTEN and UNLISTEN paths. On every (re)connect, however, handle_connect/1 replays all registered channels at once by concatenating their LISTEN statements and wrapping them in a dollar-quoted anonymous code block (DO $$BEGIN ... END$$). quote_channel/1 does not escape the $$ dollar-quote delimiter that opens and closes this block.

The listen/3 guards only reject null bytes and names longer than 63 bytes, so a channel name containing $$ passes validation unchanged. Once such a name is embedded, its $$ prematurely terminates the outer dollar-quoted string and PostgreSQL parses the remainder as additional top-level statements. Because handle_connect/1 runs on every (re)connect, the malformed replay query is rejected each time and the notification connection never re-establishes its subscriptions, silently dropping notifications for every channel sharing that connection.

An application is affected when it passes untrusted input (for example a tenant or user identifier) as a channel name to Postgrex.Notifications.listen/3. The double-quote doubling prevents forming a fully valid injected statement, so arbitrary SQL execution is not possible, but the corrupted query reliably breaks the shared notification connection for all tenants, resulting in denial of service.

This issue affects postgrex: from 0.16.0 before 0.22.3.

Workarounds

Validate channel names before passing untrusted input to Postgrex.Notifications.listen/3. Reject any name that contains the dollar-quote delimiter ($$), or restrict channel names to a safe character set such as alphanumeric characters and underscores.

Summary

SQL injection in Postgrex.Notifications.listen/3: the channel argument is interpolated straight into LISTEN "..." / UNLISTEN "..." without escaping the " character. Any caller that lets a user influence the channel name (e.g. a pub/sub bridge that uses a tenant id or topic slug as the channel) and the name is not sanitized can execute arbitrary SQL on the notifications connection.

Only those using the Postgrex.Notifications directly or via a dependency, with a non-sanitized channel name are vulnerable. Ecto does not use Postgrex.Notifications by default, but you must validate if any other dependency does.

Details

PostgreSQL escapes a " inside a quoted identifier by doubling it to "". Postgrex doesn't do that doubling, so a " inside channel closes the identifier early and everything after it is parsed as SQL on the same connection.

The notifications connection runs as whatever DB role the app is configured with. Unlike the Postgrex.query/4 API, there's no extended-protocol guard here: LISTEN/UNLISTEN are sent as simple queries, so multi-statement payloads work and the attacker can chain ; CREATE TABLE …, ; DROP …, ; CREATE ROLE …, etc.

Impact

SQL injection on the notifications connection. Affects any application that calls Postgrex.Notifications.listen/3 (or unlisten/3) with a channel name derived from untrusted input. Executes as the configured DB user with no protocol-level limit on what can be chained, so the realistic blast radius is read/modify/destroy on any data the app's DB role can reach.

Fix

Upgrade to latest Postgrex v0.22.2 or later. Altenratively, sanitize any user input given as channel name by making sure it doesn't include quotes as well as null bytes (not strictly required, but recommended):

if String.contains?(channel_name, ["\"", "\0"]) do
  raise "bad channel name"
end

Summary

SQL Injection vulnerability in elixir-ecto postgrex allows an attacker who can influence a LISTEN channel name to inject SQL into the reconnect replay query, causing a denial of service of the notification connection.

Postgrex.Notifications sanitizes channel names with quote_channel/1, which doubles double quotes so the name is safe inside a double-quoted identifier. This protects the single-statement LISTEN and UNLISTEN paths. On every (re)connect, however, handle_connect/1 replays all registered channels at once by concatenating their LISTEN statements and wrapping them in a dollar-quoted anonymous code block (DO $$BEGIN ... END$$). quote_channel/1 does not escape the $$ dollar-quote delimiter that opens and closes this block.

The listen/3 guards only reject null bytes and names longer than 63 bytes, so a channel name containing $$ passes validation unchanged. Once such a name is embedded, its $$ prematurely terminates the outer dollar-quoted string and PostgreSQL parses the remainder as additional top-level statements. Because handle_connect/1 runs on every (re)connect, the malformed replay query is rejected each time and the notification connection never re-establishes its subscriptions, silently dropping notifications for every channel sharing that connection.

An application is affected when it passes untrusted input (for example a tenant or user identifier) as a channel name to Postgrex.Notifications.listen/3. The double-quote doubling prevents forming a fully valid injected statement, so arbitrary SQL execution is not possible, but the corrupted query reliably breaks the shared notification connection for all tenants, resulting in denial of service.

This issue affects postgrex: from 0.16.0 before 0.22.3.

Workarounds

Validate channel names before passing untrusted input to Postgrex.Notifications.listen/3. Reject any name that contains the dollar-quote delimiter ($$), or restrict channel names to a safe character set such as alphanumeric characters and underscores.

Summary

SQL injection in Postgrex.Notifications.listen/3: the channel argument is interpolated straight into LISTEN "..." / UNLISTEN "..." without escaping the " character. Any caller that lets a user influence the channel name (e.g. a pub/sub bridge that uses a tenant id or topic slug as the channel) and the name is not sanitized can execute arbitrary SQL on the notifications connection.

Only those using the Postgrex.Notifications directly or via a dependency, with a non-sanitized channel name are vulnerable. Ecto does not use Postgrex.Notifications by default, but you must validate if any other dependency does.

Details

PostgreSQL escapes a " inside a quoted identifier by doubling it to "". Postgrex doesn't do that doubling, so a " inside channel closes the identifier early and everything after it is parsed as SQL on the same connection.

The notifications connection runs as whatever DB role the app is configured with. Unlike the Postgrex.query/4 API, there's no extended-protocol guard here: LISTEN/UNLISTEN are sent as simple queries, so multi-statement payloads work and the attacker can chain ; CREATE TABLE …, ; DROP …, ; CREATE ROLE …, etc.

Impact

SQL injection on the notifications connection. Affects any application that calls Postgrex.Notifications.listen/3 (or unlisten/3) with a channel name derived from untrusted input. Executes as the configured DB user with no protocol-level limit on what can be chained, so the realistic blast radius is read/modify/destroy on any data the app's DB role can reach.

Fix

Upgrade to latest Postgrex v0.22.2 or later. Altenratively, sanitize any user input given as channel name by making sure it doesn't include quotes as well as null bytes (not strictly required, but recommended):

if String.contains?(channel_name, ["\"", "\0"]) do
  raise "bad channel name"
end

Summary

SQL Injection vulnerability in elixir-ecto postgrex allows an attacker who can influence a LISTEN channel name to inject SQL into the reconnect replay query, causing a denial of service of the notification connection.

Postgrex.Notifications sanitizes channel names with quote_channel/1, which doubles double quotes so the name is safe inside a double-quoted identifier. This protects the single-statement LISTEN and UNLISTEN paths. On every (re)connect, however, handle_connect/1 replays all registered channels at once by concatenating their LISTEN statements and wrapping them in a dollar-quoted anonymous code block (DO $$BEGIN ... END$$). quote_channel/1 does not escape the $$ dollar-quote delimiter that opens and closes this block.

The listen/3 guards only reject null bytes and names longer than 63 bytes, so a channel name containing $$ passes validation unchanged. Once such a name is embedded, its $$ prematurely terminates the outer dollar-quoted string and PostgreSQL parses the remainder as additional top-level statements. Because handle_connect/1 runs on every (re)connect, the malformed replay query is rejected each time and the notification connection never re-establishes its subscriptions, silently dropping notifications for every channel sharing that connection.

An application is affected when it passes untrusted input (for example a tenant or user identifier) as a channel name to Postgrex.Notifications.listen/3. The double-quote doubling prevents forming a fully valid injected statement, so arbitrary SQL execution is not possible, but the corrupted query reliably breaks the shared notification connection for all tenants, resulting in denial of service.

This issue affects postgrex: from 0.16.0 before 0.22.3.

Workarounds

Validate channel names before passing untrusted input to Postgrex.Notifications.listen/3. Reject any name that contains the dollar-quote delimiter ($$), or restrict channel names to a safe character set such as alphanumeric characters and underscores.

Summary

SQL injection in Postgrex.Notifications.listen/3: the channel argument is interpolated straight into LISTEN "..." / UNLISTEN "..." without escaping the " character. Any caller that lets a user influence the channel name (e.g. a pub/sub bridge that uses a tenant id or topic slug as the channel) and the name is not sanitized can execute arbitrary SQL on the notifications connection.

Only those using the Postgrex.Notifications directly or via a dependency, with a non-sanitized channel name are vulnerable. Ecto does not use Postgrex.Notifications by default, but you must validate if any other dependency does.

Details

PostgreSQL escapes a " inside a quoted identifier by doubling it to "". Postgrex doesn't do that doubling, so a " inside channel closes the identifier early and everything after it is parsed as SQL on the same connection.

The notifications connection runs as whatever DB role the app is configured with. Unlike the Postgrex.query/4 API, there's no extended-protocol guard here: LISTEN/UNLISTEN are sent as simple queries, so multi-statement payloads work and the attacker can chain ; CREATE TABLE …, ; DROP …, ; CREATE ROLE …, etc.

Impact

SQL injection on the notifications connection. Affects any application that calls Postgrex.Notifications.listen/3 (or unlisten/3) with a channel name derived from untrusted input. Executes as the configured DB user with no protocol-level limit on what can be chained, so the realistic blast radius is read/modify/destroy on any data the app's DB role can reach.

Fix

Upgrade to latest Postgrex v0.22.2 or later. Altenratively, sanitize any user input given as channel name by making sure it doesn't include quotes as well as null bytes (not strictly required, but recommended):

if String.contains?(channel_name, ["\"", "\0"]) do
  raise "bad channel name"
end

Summary

SQL Injection vulnerability in elixir-ecto postgrex allows an attacker who can influence a LISTEN channel name to inject SQL into the reconnect replay query, causing a denial of service of the notification connection.

Postgrex.Notifications sanitizes channel names with quote_channel/1, which doubles double quotes so the name is safe inside a double-quoted identifier. This protects the single-statement LISTEN and UNLISTEN paths. On every (re)connect, however, handle_connect/1 replays all registered channels at once by concatenating their LISTEN statements and wrapping them in a dollar-quoted anonymous code block (DO $$BEGIN ... END$$). quote_channel/1 does not escape the $$ dollar-quote delimiter that opens and closes this block.

The listen/3 guards only reject null bytes and names longer than 63 bytes, so a channel name containing $$ passes validation unchanged. Once such a name is embedded, its $$ prematurely terminates the outer dollar-quoted string and PostgreSQL parses the remainder as additional top-level statements. Because handle_connect/1 runs on every (re)connect, the malformed replay query is rejected each time and the notification connection never re-establishes its subscriptions, silently dropping notifications for every channel sharing that connection.

An application is affected when it passes untrusted input (for example a tenant or user identifier) as a channel name to Postgrex.Notifications.listen/3. The double-quote doubling prevents forming a fully valid injected statement, so arbitrary SQL execution is not possible, but the corrupted query reliably breaks the shared notification connection for all tenants, resulting in denial of service.

This issue affects postgrex: from 0.16.0 before 0.22.3.

Workarounds

Validate channel names before passing untrusted input to Postgrex.Notifications.listen/3. Reject any name that contains the dollar-quote delimiter ($$), or restrict channel names to a safe character set such as alphanumeric characters and underscores.

Summary

SQL injection in Postgrex.Notifications.listen/3: the channel argument is interpolated straight into LISTEN "..." / UNLISTEN "..." without escaping the " character. Any caller that lets a user influence the channel name (e.g. a pub/sub bridge that uses a tenant id or topic slug as the channel) and the name is not sanitized can execute arbitrary SQL on the notifications connection.

Only those using the Postgrex.Notifications directly or via a dependency, with a non-sanitized channel name are vulnerable. Ecto does not use Postgrex.Notifications by default, but you must validate if any other dependency does.

Details

PostgreSQL escapes a " inside a quoted identifier by doubling it to "". Postgrex doesn't do that doubling, so a " inside channel closes the identifier early and everything after it is parsed as SQL on the same connection.

The notifications connection runs as whatever DB role the app is configured with. Unlike the Postgrex.query/4 API, there's no extended-protocol guard here: LISTEN/UNLISTEN are sent as simple queries, so multi-statement payloads work and the attacker can chain ; CREATE TABLE …, ; DROP …, ; CREATE ROLE …, etc.

Impact

SQL injection on the notifications connection. Affects any application that calls Postgrex.Notifications.listen/3 (or unlisten/3) with a channel name derived from untrusted input. Executes as the configured DB user with no protocol-level limit on what can be chained, so the realistic blast radius is read/modify/destroy on any data the app's DB role can reach.

Fix

Upgrade to latest Postgrex v0.22.2 or later. Altenratively, sanitize any user input given as channel name by making sure it doesn't include quotes as well as null bytes (not strictly required, but recommended):

if String.contains?(channel_name, ["\"", "\0"]) do
  raise "bad channel name"
end

Summary

SQL Injection vulnerability in elixir-ecto postgrex allows an attacker who can influence a LISTEN channel name to inject SQL into the reconnect replay query, causing a denial of service of the notification connection.

Postgrex.Notifications sanitizes channel names with quote_channel/1, which doubles double quotes so the name is safe inside a double-quoted identifier. This protects the single-statement LISTEN and UNLISTEN paths. On every (re)connect, however, handle_connect/1 replays all registered channels at once by concatenating their LISTEN statements and wrapping them in a dollar-quoted anonymous code block (DO $$BEGIN ... END$$). quote_channel/1 does not escape the $$ dollar-quote delimiter that opens and closes this block.

The listen/3 guards only reject null bytes and names longer than 63 bytes, so a channel name containing $$ passes validation unchanged. Once such a name is embedded, its $$ prematurely terminates the outer dollar-quoted string and PostgreSQL parses the remainder as additional top-level statements. Because handle_connect/1 runs on every (re)connect, the malformed replay query is rejected each time and the notification connection never re-establishes its subscriptions, silently dropping notifications for every channel sharing that connection.

An application is affected when it passes untrusted input (for example a tenant or user identifier) as a channel name to Postgrex.Notifications.listen/3. The double-quote doubling prevents forming a fully valid injected statement, so arbitrary SQL execution is not possible, but the corrupted query reliably breaks the shared notification connection for all tenants, resulting in denial of service.

This issue affects postgrex: from 0.16.0 before 0.22.3.

Workarounds

Validate channel names before passing untrusted input to Postgrex.Notifications.listen/3. Reject any name that contains the dollar-quote delimiter ($$), or restrict channel names to a safe character set such as alphanumeric characters and underscores.

Summary

SQL injection in Postgrex.Notifications.listen/3: the channel argument is interpolated straight into LISTEN "..." / UNLISTEN "..." without escaping the " character. Any caller that lets a user influence the channel name (e.g. a pub/sub bridge that uses a tenant id or topic slug as the channel) and the name is not sanitized can execute arbitrary SQL on the notifications connection.

Only those using the Postgrex.Notifications directly or via a dependency, with a non-sanitized channel name are vulnerable. Ecto does not use Postgrex.Notifications by default, but you must validate if any other dependency does.

Details

PostgreSQL escapes a " inside a quoted identifier by doubling it to "". Postgrex doesn't do that doubling, so a " inside channel closes the identifier early and everything after it is parsed as SQL on the same connection.

The notifications connection runs as whatever DB role the app is configured with. Unlike the Postgrex.query/4 API, there's no extended-protocol guard here: LISTEN/UNLISTEN are sent as simple queries, so multi-statement payloads work and the attacker can chain ; CREATE TABLE …, ; DROP …, ; CREATE ROLE …, etc.

Impact

SQL injection on the notifications connection. Affects any application that calls Postgrex.Notifications.listen/3 (or unlisten/3) with a channel name derived from untrusted input. Executes as the configured DB user with no protocol-level limit on what can be chained, so the realistic blast radius is read/modify/destroy on any data the app's DB role can reach.

Fix

Upgrade to latest Postgrex v0.22.2 or later. Altenratively, sanitize any user input given as channel name by making sure it doesn't include quotes as well as null bytes (not strictly required, but recommended):

if String.contains?(channel_name, ["\"", "\0"]) do
  raise "bad channel name"
end

Summary

SQL Injection vulnerability in elixir-ecto postgrex allows an attacker who can influence a LISTEN channel name to inject SQL into the reconnect replay query, causing a denial of service of the notification connection.

Postgrex.Notifications sanitizes channel names with quote_channel/1, which doubles double quotes so the name is safe inside a double-quoted identifier. This protects the single-statement LISTEN and UNLISTEN paths. On every (re)connect, however, handle_connect/1 replays all registered channels at once by concatenating their LISTEN statements and wrapping them in a dollar-quoted anonymous code block (DO $$BEGIN ... END$$). quote_channel/1 does not escape the $$ dollar-quote delimiter that opens and closes this block.

The listen/3 guards only reject null bytes and names longer than 63 bytes, so a channel name containing $$ passes validation unchanged. Once such a name is embedded, its $$ prematurely terminates the outer dollar-quoted string and PostgreSQL parses the remainder as additional top-level statements. Because handle_connect/1 runs on every (re)connect, the malformed replay query is rejected each time and the notification connection never re-establishes its subscriptions, silently dropping notifications for every channel sharing that connection.

An application is affected when it passes untrusted input (for example a tenant or user identifier) as a channel name to Postgrex.Notifications.listen/3. The double-quote doubling prevents forming a fully valid injected statement, so arbitrary SQL execution is not possible, but the corrupted query reliably breaks the shared notification connection for all tenants, resulting in denial of service.

This issue affects postgrex: from 0.16.0 before 0.22.3.

Workarounds

Validate channel names before passing untrusted input to Postgrex.Notifications.listen/3. Reject any name that contains the dollar-quote delimiter ($$), or restrict channel names to a safe character set such as alphanumeric characters and underscores.

Summary

SQL injection in Postgrex.Notifications.listen/3: the channel argument is interpolated straight into LISTEN "..." / UNLISTEN "..." without escaping the " character. Any caller that lets a user influence the channel name (e.g. a pub/sub bridge that uses a tenant id or topic slug as the channel) and the name is not sanitized can execute arbitrary SQL on the notifications connection.

Only those using the Postgrex.Notifications directly or via a dependency, with a non-sanitized channel name are vulnerable. Ecto does not use Postgrex.Notifications by default, but you must validate if any other dependency does.

Details

PostgreSQL escapes a " inside a quoted identifier by doubling it to "". Postgrex doesn't do that doubling, so a " inside channel closes the identifier early and everything after it is parsed as SQL on the same connection.

The notifications connection runs as whatever DB role the app is configured with. Unlike the Postgrex.query/4 API, there's no extended-protocol guard here: LISTEN/UNLISTEN are sent as simple queries, so multi-statement payloads work and the attacker can chain ; CREATE TABLE …, ; DROP …, ; CREATE ROLE …, etc.

Impact

SQL injection on the notifications connection. Affects any application that calls Postgrex.Notifications.listen/3 (or unlisten/3) with a channel name derived from untrusted input. Executes as the configured DB user with no protocol-level limit on what can be chained, so the realistic blast radius is read/modify/destroy on any data the app's DB role can reach.

Fix

Upgrade to latest Postgrex v0.22.2 or later. Altenratively, sanitize any user input given as channel name by making sure it doesn't include quotes as well as null bytes (not strictly required, but recommended):

if String.contains?(channel_name, ["\"", "\0"]) do
  raise "bad channel name"
end

Summary

SQL Injection vulnerability in elixir-ecto postgrex allows an attacker who can influence a LISTEN channel name to inject SQL into the reconnect replay query, causing a denial of service of the notification connection.

Postgrex.Notifications sanitizes channel names with quote_channel/1, which doubles double quotes so the name is safe inside a double-quoted identifier. This protects the single-statement LISTEN and UNLISTEN paths. On every (re)connect, however, handle_connect/1 replays all registered channels at once by concatenating their LISTEN statements and wrapping them in a dollar-quoted anonymous code block (DO $$BEGIN ... END$$). quote_channel/1 does not escape the $$ dollar-quote delimiter that opens and closes this block.

The listen/3 guards only reject null bytes and names longer than 63 bytes, so a channel name containing $$ passes validation unchanged. Once such a name is embedded, its $$ prematurely terminates the outer dollar-quoted string and PostgreSQL parses the remainder as additional top-level statements. Because handle_connect/1 runs on every (re)connect, the malformed replay query is rejected each time and the notification connection never re-establishes its subscriptions, silently dropping notifications for every channel sharing that connection.

An application is affected when it passes untrusted input (for example a tenant or user identifier) as a channel name to Postgrex.Notifications.listen/3. The double-quote doubling prevents forming a fully valid injected statement, so arbitrary SQL execution is not possible, but the corrupted query reliably breaks the shared notification connection for all tenants, resulting in denial of service.

This issue affects postgrex: from 0.16.0 before 0.22.3.

Workarounds

Validate channel names before passing untrusted input to Postgrex.Notifications.listen/3. Reject any name that contains the dollar-quote delimiter ($$), or restrict channel names to a safe character set such as alphanumeric characters and underscores.

Summary

SQL injection in Postgrex.Notifications.listen/3: the channel argument is interpolated straight into LISTEN "..." / UNLISTEN "..." without escaping the " character. Any caller that lets a user influence the channel name (e.g. a pub/sub bridge that uses a tenant id or topic slug as the channel) and the name is not sanitized can execute arbitrary SQL on the notifications connection.

Only those using the Postgrex.Notifications directly or via a dependency, with a non-sanitized channel name are vulnerable. Ecto does not use Postgrex.Notifications by default, but you must validate if any other dependency does.

Details

PostgreSQL escapes a " inside a quoted identifier by doubling it to "". Postgrex doesn't do that doubling, so a " inside channel closes the identifier early and everything after it is parsed as SQL on the same connection.

The notifications connection runs as whatever DB role the app is configured with. Unlike the Postgrex.query/4 API, there's no extended-protocol guard here: LISTEN/UNLISTEN are sent as simple queries, so multi-statement payloads work and the attacker can chain ; CREATE TABLE …, ; DROP …, ; CREATE ROLE …, etc.

Impact

SQL injection on the notifications connection. Affects any application that calls Postgrex.Notifications.listen/3 (or unlisten/3) with a channel name derived from untrusted input. Executes as the configured DB user with no protocol-level limit on what can be chained, so the realistic blast radius is read/modify/destroy on any data the app's DB role can reach.

Fix

Upgrade to latest Postgrex v0.22.2 or later. Altenratively, sanitize any user input given as channel name by making sure it doesn't include quotes as well as null bytes (not strictly required, but recommended):

if String.contains?(channel_name, ["\"", "\0"]) do
  raise "bad channel name"
end

Summary

SQL Injection vulnerability in elixir-ecto postgrex allows an attacker who can influence a LISTEN channel name to inject SQL into the reconnect replay query, causing a denial of service of the notification connection.

Postgrex.Notifications sanitizes channel names with quote_channel/1, which doubles double quotes so the name is safe inside a double-quoted identifier. This protects the single-statement LISTEN and UNLISTEN paths. On every (re)connect, however, handle_connect/1 replays all registered channels at once by concatenating their LISTEN statements and wrapping them in a dollar-quoted anonymous code block (DO $$BEGIN ... END$$). quote_channel/1 does not escape the $$ dollar-quote delimiter that opens and closes this block.

The listen/3 guards only reject null bytes and names longer than 63 bytes, so a channel name containing $$ passes validation unchanged. Once such a name is embedded, its $$ prematurely terminates the outer dollar-quoted string and PostgreSQL parses the remainder as additional top-level statements. Because handle_connect/1 runs on every (re)connect, the malformed replay query is rejected each time and the notification connection never re-establishes its subscriptions, silently dropping notifications for every channel sharing that connection.

An application is affected when it passes untrusted input (for example a tenant or user identifier) as a channel name to Postgrex.Notifications.listen/3. The double-quote doubling prevents forming a fully valid injected statement, so arbitrary SQL execution is not possible, but the corrupted query reliably breaks the shared notification connection for all tenants, resulting in denial of service.

This issue affects postgrex: from 0.16.0 before 0.22.3.

Workarounds

Validate channel names before passing untrusted input to Postgrex.Notifications.listen/3. Reject any name that contains the dollar-quote delimiter ($$), or restrict channel names to a safe character set such as alphanumeric characters and underscores.

Summary

SQL injection in Postgrex.Notifications.listen/3: the channel argument is interpolated straight into LISTEN "..." / UNLISTEN "..." without escaping the " character. Any caller that lets a user influence the channel name (e.g. a pub/sub bridge that uses a tenant id or topic slug as the channel) and the name is not sanitized can execute arbitrary SQL on the notifications connection.

Only those using the Postgrex.Notifications directly or via a dependency, with a non-sanitized channel name are vulnerable. Ecto does not use Postgrex.Notifications by default, but you must validate if any other dependency does.

Details

PostgreSQL escapes a " inside a quoted identifier by doubling it to "". Postgrex doesn't do that doubling, so a " inside channel closes the identifier early and everything after it is parsed as SQL on the same connection.

The notifications connection runs as whatever DB role the app is configured with. Unlike the Postgrex.query/4 API, there's no extended-protocol guard here: LISTEN/UNLISTEN are sent as simple queries, so multi-statement payloads work and the attacker can chain ; CREATE TABLE …, ; DROP …, ; CREATE ROLE …, etc.

Impact

SQL injection on the notifications connection. Affects any application that calls Postgrex.Notifications.listen/3 (or unlisten/3) with a channel name derived from untrusted input. Executes as the configured DB user with no protocol-level limit on what can be chained, so the realistic blast radius is read/modify/destroy on any data the app's DB role can reach.

Fix

Upgrade to latest Postgrex v0.22.2 or later. Altenratively, sanitize any user input given as channel name by making sure it doesn't include quotes as well as null bytes (not strictly required, but recommended):

if String.contains?(channel_name, ["\"", "\0"]) do
  raise "bad channel name"
end

Summary

SQL Injection vulnerability in elixir-ecto postgrex allows an attacker who can influence a LISTEN channel name to inject SQL into the reconnect replay query, causing a denial of service of the notification connection.

Postgrex.Notifications sanitizes channel names with quote_channel/1, which doubles double quotes so the name is safe inside a double-quoted identifier. This protects the single-statement LISTEN and UNLISTEN paths. On every (re)connect, however, handle_connect/1 replays all registered channels at once by concatenating their LISTEN statements and wrapping them in a dollar-quoted anonymous code block (DO $$BEGIN ... END$$). quote_channel/1 does not escape the $$ dollar-quote delimiter that opens and closes this block.

The listen/3 guards only reject null bytes and names longer than 63 bytes, so a channel name containing $$ passes validation unchanged. Once such a name is embedded, its $$ prematurely terminates the outer dollar-quoted string and PostgreSQL parses the remainder as additional top-level statements. Because handle_connect/1 runs on every (re)connect, the malformed replay query is rejected each time and the notification connection never re-establishes its subscriptions, silently dropping notifications for every channel sharing that connection.

An application is affected when it passes untrusted input (for example a tenant or user identifier) as a channel name to Postgrex.Notifications.listen/3. The double-quote doubling prevents forming a fully valid injected statement, so arbitrary SQL execution is not possible, but the corrupted query reliably breaks the shared notification connection for all tenants, resulting in denial of service.

This issue affects postgrex: from 0.16.0 before 0.22.3.

Workarounds

Validate channel names before passing untrusted input to Postgrex.Notifications.listen/3. Reject any name that contains the dollar-quote delimiter ($$), or restrict channel names to a safe character set such as alphanumeric characters and underscores.

Summary

SQL injection in Postgrex.Notifications.listen/3: the channel argument is interpolated straight into LISTEN "..." / UNLISTEN "..." without escaping the " character. Any caller that lets a user influence the channel name (e.g. a pub/sub bridge that uses a tenant id or topic slug as the channel) and the name is not sanitized can execute arbitrary SQL on the notifications connection.

Only those using the Postgrex.Notifications directly or via a dependency, with a non-sanitized channel name are vulnerable. Ecto does not use Postgrex.Notifications by default, but you must validate if any other dependency does.

Details

PostgreSQL escapes a " inside a quoted identifier by doubling it to "". Postgrex doesn't do that doubling, so a " inside channel closes the identifier early and everything after it is parsed as SQL on the same connection.

The notifications connection runs as whatever DB role the app is configured with. Unlike the Postgrex.query/4 API, there's no extended-protocol guard here: LISTEN/UNLISTEN are sent as simple queries, so multi-statement payloads work and the attacker can chain ; CREATE TABLE …, ; DROP …, ; CREATE ROLE …, etc.

Impact

SQL injection on the notifications connection. Affects any application that calls Postgrex.Notifications.listen/3 (or unlisten/3) with a channel name derived from untrusted input. Executes as the configured DB user with no protocol-level limit on what can be chained, so the realistic blast radius is read/modify/destroy on any data the app's DB role can reach.

Fix

Upgrade to latest Postgrex v0.22.2 or later. Altenratively, sanitize any user input given as channel name by making sure it doesn't include quotes as well as null bytes (not strictly required, but recommended):

if String.contains?(channel_name, ["\"", "\0"]) do
  raise "bad channel name"
end

Summary

SQL Injection vulnerability in elixir-ecto postgrex allows an attacker who can influence a LISTEN channel name to inject SQL into the reconnect replay query, causing a denial of service of the notification connection.

Postgrex.Notifications sanitizes channel names with quote_channel/1, which doubles double quotes so the name is safe inside a double-quoted identifier. This protects the single-statement LISTEN and UNLISTEN paths. On every (re)connect, however, handle_connect/1 replays all registered channels at once by concatenating their LISTEN statements and wrapping them in a dollar-quoted anonymous code block (DO $$BEGIN ... END$$). quote_channel/1 does not escape the $$ dollar-quote delimiter that opens and closes this block.

The listen/3 guards only reject null bytes and names longer than 63 bytes, so a channel name containing $$ passes validation unchanged. Once such a name is embedded, its $$ prematurely terminates the outer dollar-quoted string and PostgreSQL parses the remainder as additional top-level statements. Because handle_connect/1 runs on every (re)connect, the malformed replay query is rejected each time and the notification connection never re-establishes its subscriptions, silently dropping notifications for every channel sharing that connection.

An application is affected when it passes untrusted input (for example a tenant or user identifier) as a channel name to Postgrex.Notifications.listen/3. The double-quote doubling prevents forming a fully valid injected statement, so arbitrary SQL execution is not possible, but the corrupted query reliably breaks the shared notification connection for all tenants, resulting in denial of service.

This issue affects postgrex: from 0.16.0 before 0.22.3.

Workarounds

Validate channel names before passing untrusted input to Postgrex.Notifications.listen/3. Reject any name that contains the dollar-quote delimiter ($$), or restrict channel names to a safe character set such as alphanumeric characters and underscores.

Summary

SQL injection in Postgrex.Notifications.listen/3: the channel argument is interpolated straight into LISTEN "..." / UNLISTEN "..." without escaping the " character. Any caller that lets a user influence the channel name (e.g. a pub/sub bridge that uses a tenant id or topic slug as the channel) and the name is not sanitized can execute arbitrary SQL on the notifications connection.

Only those using the Postgrex.Notifications directly or via a dependency, with a non-sanitized channel name are vulnerable. Ecto does not use Postgrex.Notifications by default, but you must validate if any other dependency does.

Details

PostgreSQL escapes a " inside a quoted identifier by doubling it to "". Postgrex doesn't do that doubling, so a " inside channel closes the identifier early and everything after it is parsed as SQL on the same connection.

The notifications connection runs as whatever DB role the app is configured with. Unlike the Postgrex.query/4 API, there's no extended-protocol guard here: LISTEN/UNLISTEN are sent as simple queries, so multi-statement payloads work and the attacker can chain ; CREATE TABLE …, ; DROP …, ; CREATE ROLE …, etc.

Impact

SQL injection on the notifications connection. Affects any application that calls Postgrex.Notifications.listen/3 (or unlisten/3) with a channel name derived from untrusted input. Executes as the configured DB user with no protocol-level limit on what can be chained, so the realistic blast radius is read/modify/destroy on any data the app's DB role can reach.

Fix

Upgrade to latest Postgrex v0.22.2 or later. Altenratively, sanitize any user input given as channel name by making sure it doesn't include quotes as well as null bytes (not strictly required, but recommended):

if String.contains?(channel_name, ["\"", "\0"]) do
  raise "bad channel name"
end

Summary

SQL Injection vulnerability in elixir-ecto postgrex allows an attacker who can influence a LISTEN channel name to inject SQL into the reconnect replay query, causing a denial of service of the notification connection.

Postgrex.Notifications sanitizes channel names with quote_channel/1, which doubles double quotes so the name is safe inside a double-quoted identifier. This protects the single-statement LISTEN and UNLISTEN paths. On every (re)connect, however, handle_connect/1 replays all registered channels at once by concatenating their LISTEN statements and wrapping them in a dollar-quoted anonymous code block (DO $$BEGIN ... END$$). quote_channel/1 does not escape the $$ dollar-quote delimiter that opens and closes this block.

The listen/3 guards only reject null bytes and names longer than 63 bytes, so a channel name containing $$ passes validation unchanged. Once such a name is embedded, its $$ prematurely terminates the outer dollar-quoted string and PostgreSQL parses the remainder as additional top-level statements. Because handle_connect/1 runs on every (re)connect, the malformed replay query is rejected each time and the notification connection never re-establishes its subscriptions, silently dropping notifications for every channel sharing that connection.

An application is affected when it passes untrusted input (for example a tenant or user identifier) as a channel name to Postgrex.Notifications.listen/3. The double-quote doubling prevents forming a fully valid injected statement, so arbitrary SQL execution is not possible, but the corrupted query reliably breaks the shared notification connection for all tenants, resulting in denial of service.

This issue affects postgrex: from 0.16.0 before 0.22.3.

Workarounds

Validate channel names before passing untrusted input to Postgrex.Notifications.listen/3. Reject any name that contains the dollar-quote delimiter ($$), or restrict channel names to a safe character set such as alphanumeric characters and underscores.

Summary

SQL injection in Postgrex.Notifications.listen/3: the channel argument is interpolated straight into LISTEN "..." / UNLISTEN "..." without escaping the " character. Any caller that lets a user influence the channel name (e.g. a pub/sub bridge that uses a tenant id or topic slug as the channel) and the name is not sanitized can execute arbitrary SQL on the notifications connection.

Only those using the Postgrex.Notifications directly or via a dependency, with a non-sanitized channel name are vulnerable. Ecto does not use Postgrex.Notifications by default, but you must validate if any other dependency does.

Details

PostgreSQL escapes a " inside a quoted identifier by doubling it to "". Postgrex doesn't do that doubling, so a " inside channel closes the identifier early and everything after it is parsed as SQL on the same connection.

The notifications connection runs as whatever DB role the app is configured with. Unlike the Postgrex.query/4 API, there's no extended-protocol guard here: LISTEN/UNLISTEN are sent as simple queries, so multi-statement payloads work and the attacker can chain ; CREATE TABLE …, ; DROP …, ; CREATE ROLE …, etc.

Impact

SQL injection on the notifications connection. Affects any application that calls Postgrex.Notifications.listen/3 (or unlisten/3) with a channel name derived from untrusted input. Executes as the configured DB user with no protocol-level limit on what can be chained, so the realistic blast radius is read/modify/destroy on any data the app's DB role can reach.

Fix

Upgrade to latest Postgrex v0.22.2 or later. Altenratively, sanitize any user input given as channel name by making sure it doesn't include quotes as well as null bytes (not strictly required, but recommended):

if String.contains?(channel_name, ["\"", "\0"]) do
  raise "bad channel name"
end

Summary

SQL Injection vulnerability in elixir-ecto postgrex allows an attacker who can influence a LISTEN channel name to inject SQL into the reconnect replay query, causing a denial of service of the notification connection.

Postgrex.Notifications sanitizes channel names with quote_channel/1, which doubles double quotes so the name is safe inside a double-quoted identifier. This protects the single-statement LISTEN and UNLISTEN paths. On every (re)connect, however, handle_connect/1 replays all registered channels at once by concatenating their LISTEN statements and wrapping them in a dollar-quoted anonymous code block (DO $$BEGIN ... END$$). quote_channel/1 does not escape the $$ dollar-quote delimiter that opens and closes this block.

The listen/3 guards only reject null bytes and names longer than 63 bytes, so a channel name containing $$ passes validation unchanged. Once such a name is embedded, its $$ prematurely terminates the outer dollar-quoted string and PostgreSQL parses the remainder as additional top-level statements. Because handle_connect/1 runs on every (re)connect, the malformed replay query is rejected each time and the notification connection never re-establishes its subscriptions, silently dropping notifications for every channel sharing that connection.

An application is affected when it passes untrusted input (for example a tenant or user identifier) as a channel name to Postgrex.Notifications.listen/3. The double-quote doubling prevents forming a fully valid injected statement, so arbitrary SQL execution is not possible, but the corrupted query reliably breaks the shared notification connection for all tenants, resulting in denial of service.

This issue affects postgrex: from 0.16.0 before 0.22.3.

Workarounds

Validate channel names before passing untrusted input to Postgrex.Notifications.listen/3. Reject any name that contains the dollar-quote delimiter ($$), or restrict channel names to a safe character set such as alphanumeric characters and underscores.

Summary

SQL injection in Postgrex.Notifications.listen/3: the channel argument is interpolated straight into LISTEN "..." / UNLISTEN "..." without escaping the " character. Any caller that lets a user influence the channel name (e.g. a pub/sub bridge that uses a tenant id or topic slug as the channel) and the name is not sanitized can execute arbitrary SQL on the notifications connection.

Only those using the Postgrex.Notifications directly or via a dependency, with a non-sanitized channel name are vulnerable. Ecto does not use Postgrex.Notifications by default, but you must validate if any other dependency does.

Details

PostgreSQL escapes a " inside a quoted identifier by doubling it to "". Postgrex doesn't do that doubling, so a " inside channel closes the identifier early and everything after it is parsed as SQL on the same connection.

The notifications connection runs as whatever DB role the app is configured with. Unlike the Postgrex.query/4 API, there's no extended-protocol guard here: LISTEN/UNLISTEN are sent as simple queries, so multi-statement payloads work and the attacker can chain ; CREATE TABLE …, ; DROP …, ; CREATE ROLE …, etc.

Impact

SQL injection on the notifications connection. Affects any application that calls Postgrex.Notifications.listen/3 (or unlisten/3) with a channel name derived from untrusted input. Executes as the configured DB user with no protocol-level limit on what can be chained, so the realistic blast radius is read/modify/destroy on any data the app's DB role can reach.

Fix

Upgrade to latest Postgrex v0.22.2 or later. Altenratively, sanitize any user input given as channel name by making sure it doesn't include quotes as well as null bytes (not strictly required, but recommended):

if String.contains?(channel_name, ["\"", "\0"]) do
  raise "bad channel name"
end